CN114117206A - Recommendation model processing method and device, electronic equipment and storage medium - Google Patents

Recommendation model processing method and device, electronic equipment and storage medium Download PDF

Info

Publication number
CN114117206A
CN114117206A CN202111320914.2A CN202111320914A CN114117206A CN 114117206 A CN114117206 A CN 114117206A CN 202111320914 A CN202111320914 A CN 202111320914A CN 114117206 A CN114117206 A CN 114117206A
Authority
CN
China
Prior art keywords
model
recommendation
operator
sample
precision
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202111320914.2A
Other languages
Chinese (zh)
Other versions
CN114117206B (en
Inventor
谈建超
陆顺
李吉祥
朱文涛
杨森
刘霁
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Dajia Internet Information Technology Co Ltd
Original Assignee
Beijing Dajia Internet Information Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Dajia Internet Information Technology Co Ltd filed Critical Beijing Dajia Internet Information Technology Co Ltd
Priority to CN202111320914.2A priority Critical patent/CN114117206B/en
Publication of CN114117206A publication Critical patent/CN114117206A/en
Application granted granted Critical
Publication of CN114117206B publication Critical patent/CN114117206B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods

Abstract

The disclosure relates to a recommendation model processing method and device, electronic equipment and a storage medium, and belongs to the technical field of computers. The method comprises the following steps: acquiring a recommendation model, wherein the recommendation model comprises a plurality of operators which are arranged in sequence, and the operators sequentially process the account data and the article data which are input into the recommendation model according to the sequence to obtain recommendation parameters; based on the recommendation model, acquiring operator characteristics corresponding to each operator in the recommendation model and structural characteristics of the recommendation model, wherein the structural characteristics represent a connection relation between any two operators in the multiple operators; extracting features based on the operator features and the structural features corresponding to each operator to obtain model features corresponding to the recommended models; and performing precision prediction based on the model characteristics to obtain recommendation precision corresponding to the recommendation model. The model features obtained by the method can more accurately represent the recommendation model, so that the recommendation precision obtained based on the model feature prediction is more accurate, and the accuracy of the recommendation precision is improved.

Description

Recommendation model processing method and device, electronic equipment and storage medium
Technical Field
The present disclosure relates to the field of computer technologies, and in particular, to a recommendation model processing method and apparatus, an electronic device, and a storage medium.
Background
With the development of computer technology, building a network model by adopting a neural architecture search mode has become a common model building method. Furthermore, after obtaining the network model, it is usually necessary to determine a model accuracy corresponding to the network model, and determine whether the network model is actually applicable according to the model accuracy. For example, in a recommendation scenario, recommendation accuracy corresponding to a recommendation model may be determined based on model features corresponding to the recommendation model, where how to obtain the model features corresponding to the recommendation model is a key for determining the recommendation accuracy.
Disclosure of Invention
The disclosure provides a recommendation model processing method, a recommendation model processing device, an electronic device and a storage medium, and accuracy of determined recommendation precision is improved.
According to an aspect of the embodiments of the present disclosure, there is provided a recommendation model processing method, including:
acquiring a recommendation model, wherein the recommendation model comprises a plurality of operators which are arranged in sequence, the operators sequentially process account data and article data which are input into the recommendation model according to the sequence to obtain recommendation parameters, and the recommendation parameters indicate whether the articles are recommended to the accounts or not;
acquiring operator characteristics corresponding to each operator in the recommendation model and structural characteristics of the recommendation model based on the recommendation model, wherein the structural characteristics represent the connection relation between any two operators in the multiple operators;
performing feature extraction based on the operator features corresponding to each operator and the structural features to obtain model features corresponding to the recommended models;
and performing precision prediction based on the model characteristics to obtain recommendation precision corresponding to the recommendation model.
According to the method provided by the embodiment of the disclosure, the operator characteristics and the structural characteristics corresponding to the recommendation model are obtained firstly, and then the effective model characteristics are extracted by utilizing the operator characteristics and the structural characteristics, so that the characteristics of the recommendation model can be fully obtained by the mode of obtaining the model characteristics, the obtained model characteristics can more accurately represent the recommendation model, the recommendation precision obtained based on the model characteristics is more accurate, and the accuracy of the recommendation precision is improved.
In some embodiments, the recommendation model further includes an index identifier corresponding to each operator, and the step of obtaining an operator feature corresponding to each operator includes:
inquiring operator characteristics corresponding to each operator from an operator characteristic table based on the index identification corresponding to each operator;
the operator feature table comprises index identifications and operator features corresponding to at least one operator, wherein the operator features are obtained by performing feature extraction on the operators.
In the embodiment of the disclosure, since the operator characteristic table already includes the operator characteristic, the operator characteristic can be quickly obtained by querying the operator characteristic table, and the operator characteristic corresponding to the operator does not need to be temporarily extracted, so that the efficiency of obtaining the recommendation precision is improved.
In some embodiments, the recommendation model includes N operators, where N is an integer greater than 1, and the step of obtaining the structural feature includes:
creating a target matrix with N rows and N columns based on the N operators in the recommendation model, wherein the N operators are respectively used as the N rows and the N columns of the target matrix, the rows and the columns corresponding to the elements on the diagonal line in the target matrix are the same operators, the value of the element on the diagonal line in the target matrix is equal to the number of other operators connected with the operators corresponding to the elements, the rows and the columns corresponding to the elements at other positions in the target matrix are two different operators, and the values of the elements at other positions indicate whether the two operators corresponding to the elements have a connection relation;
and extracting the characteristics of the target matrix to obtain the structural characteristics.
In the embodiment of the disclosure, the target matrix can fully represent the connection relationship between any two operators in the recommendation model, that is, the topological structure information of the recommendation model is fully obtained, so that the structural features obtained based on the target matrix can accurately represent the structure of the recommendation model, and the accuracy of the structural features is improved.
In some embodiments, the performing feature extraction on the operator feature and the structural feature of each operator to obtain a model feature corresponding to the recommended model includes:
splicing the operator characteristics corresponding to each operator and the structural characteristics to obtain splicing characteristics corresponding to the recommended model;
and coding the splicing characteristics to obtain the model characteristics.
In the embodiment of the disclosure, the plurality of operator features and the structural features are respectively obtained, the obtained operator features can accurately represent operators in the recommendation model, and the structural features can accurately represent the structure of the recommendation model, so that the model features obtained by splicing and encoding the plurality of operator features and the structural features can accurately represent the recommendation model, and the accuracy of the model features is improved.
In some embodiments, the recommendation precision prediction model includes a feature extraction network and a prediction network,
the feature extraction network is used for performing feature extraction based on the operator features corresponding to each operator and the structural features to obtain the model features;
and the prediction network is used for predicting the accuracy of the model characteristics to obtain the recommendation accuracy.
In some embodiments, the recommendation accuracy prediction model further comprises a feature acquisition network,
the feature acquisition network is used for acquiring operator features corresponding to each operator in the recommendation model;
the feature acquisition network is further configured to acquire the structural features of the recommendation model.
In the embodiment of the disclosure, compared with a mode that a recommendation model is trained in the related art, then the recommendation model is tested by adopting test data, and the recommendation precision is determined according to a test result, the recommendation precision is predicted by utilizing a recommendation precision prediction model, so that the recommendation precision can be rapidly obtained, and the efficiency of obtaining the recommendation precision is improved.
In some embodiments, the training process of the recommendation accuracy prediction model comprises:
obtaining a sample operator characteristic, a first sample structural characteristic and a first sample recommendation precision corresponding to each sample operator in a first sample recommendation model, wherein the first sample structural characteristic represents a connection relation between any two sample operators in the first sample recommendation model;
calling the recommendation precision prediction model, and processing the sample operator characteristics and the first sample structure characteristics corresponding to each sample operator in the first sample recommendation model to obtain first prediction recommendation precision;
training the recommendation precision prediction model based on the first prediction recommendation precision and the first sample recommendation precision.
In some embodiments, the recommendation model processing method further comprises:
inputting the account data and the item data into the recommendation model;
calling a first operator in the recommendation model, and processing the account data and the article data to obtain output data of the first operator;
and taking the output data of the first operator as the input data of the next operator connected with the first operator, calling the next operator to process the output data until the output data of the last operator in the recommendation model is obtained, and determining the output data of the last operator as the recommendation parameter.
In some embodiments, the obtaining a recommendation model includes:
acquiring a plurality of recommendation models, wherein different operators exist in any two recommendation models, or the connection relation between the operators in any two recommendation models is different;
after the accuracy prediction is performed based on the model features and the recommendation accuracy corresponding to the recommendation model is obtained, the recommendation model processing method further includes:
selecting a target recommendation model corresponding to the maximum recommendation precision from the plurality of recommendation models based on the recommendation precisions corresponding to the plurality of recommendation models;
and calling the target recommendation model for recommendation.
In the embodiment of the disclosure, based on the recommendation precision corresponding to the plurality of recommendation models, the most accurate recommendation model can be selected from the plurality of recommendation models, so that when the selected recommendation model is applied to an actual recommendation scene, accurate recommendation can be performed based on the recommendation model, and the recommendation accuracy is improved.
In some embodiments, the recommendation model further includes a plurality of edges, each edge connects two operators, and the operators are configured to process input data input to the operators to obtain output data; and the output data of one operator connected by each edge is used as the input data of the other operator connected by each edge.
According to another aspect of the embodiments of the present disclosure, there is provided a recommendation accuracy prediction model training method, including:
in some embodiments, the recommendation model processing method further comprises:
obtaining a sample operator characteristic, a first sample structural characteristic and a first sample recommendation precision corresponding to each sample operator in a first sample recommendation model, wherein the first sample structural characteristic represents a connection relation between any two sample operators in the first sample recommendation model;
calling a recommendation precision prediction model, performing feature extraction on sample operator features and first sample structure features corresponding to each sample operator in the first sample recommendation model to obtain model features corresponding to the first sample recommendation model, and performing precision prediction based on the model features corresponding to the first sample recommendation model to obtain first prediction recommendation precision;
training the recommendation precision prediction model based on the first prediction recommendation precision and the first sample recommendation precision.
According to the method provided by the embodiment of the disclosure, the recommendation precision prediction model for predicting the recommendation precision is trained, then, the recommendation precision prediction model can be utilized to process each operator characteristic and structural characteristic corresponding to the recommendation model, and effective model characteristics are extracted by utilizing the operator characteristics and the structural characteristics, so that the characteristics of the recommendation model are fully obtained, the obtained model characteristics can more accurately represent the recommendation model, the recommendation precision obtained by the recommendation precision prediction model based on the model characteristics is more accurate, and the accuracy of the recommendation precision is improved.
In some embodiments, the recommendation accuracy prediction model training method further comprises:
obtaining sample operator characteristics, second sample structure characteristics and proxy recommendation precision corresponding to each sample operator in a second sample recommendation model, wherein the second sample structure characteristics represent a connection relation between any two sample operators in the second sample recommendation model, and the proxy recommendation precision is a variable for indicating the recommendation precision of the second sample recommendation model;
calling the recommendation precision prediction model, performing feature extraction on the sample operator features and the second sample structure features corresponding to each sample operator in the second sample recommendation model to obtain model features corresponding to the second sample recommendation model, and performing precision prediction based on the model features corresponding to the second sample recommendation model to obtain second prediction recommendation precision;
the training the recommendation accuracy prediction model based on the first prediction recommendation accuracy and the first sample recommendation accuracy includes:
and training the recommendation precision prediction model based on the first prediction recommendation precision, the first sample recommendation precision, the second prediction recommendation precision and the agent recommendation precision.
In the embodiment of the disclosure, when the recommendation precision prediction model is trained, not only the first sample recommendation model is adopted to train the recommendation precision prediction model, but also the second sample recommendation model is adopted to train the recommendation precision prediction model, so that more sample models for training are provided, and the accuracy and generalization capability of the recommendation precision prediction model obtained by training are improved.
In some embodiments, said training said recommendation accuracy prediction model based on said first predicted recommendation accuracy, said first sample recommendation accuracy, said second predicted recommendation accuracy, and said agent recommendation accuracy comprises:
acquiring a first error between the first prediction recommendation precision and the first sample recommendation precision, and acquiring a second error between the second prediction recommendation precision and the second sample recommendation precision;
and adjusting model parameters in the recommendation precision prediction model and the proxy recommendation precision based on the first error and the second error so that the error between the prediction recommendation precision corresponding to the second sample recommendation model at the historical time and the adjusted proxy recommendation precision approaches to a target error corresponding to the historical time.
In some embodiments, the recommendation accuracy prediction model is trained based on the first prediction recommendation accuracy, the first sample recommendation accuracy, the second prediction recommendation accuracy, and the proxy recommendation accuracy using the following penalty function:
Figure BDA0003345560400000051
wherein L represents the loss function, fθ(. cndot.) represents the recommendation accuracy prediction model, θ represents a model parameter in the recommendation accuracy prediction model, fθ(xi) Represents the first prediction recommendation precision, x, corresponding to the ith first sample recommendation modeliRepresenting the sample operator characteristic and the corresponding sample operator characteristic of each sample operator in the ith first sample recommendation modelThe first sample structure characteristic, n represents the number of the first sample recommendation models used for training the recommendation precision prediction model, yiRepresenting the first sample recommendation precision, f, corresponding to the ith first sample recommendation modelθ(vj) Represents the second prediction recommendation precision, v, corresponding to the jth second sample recommendation modeljRepresenting sample operator characteristics and the second sample structure characteristics corresponding to each sample operator in the jth second sample recommendation model, V representing the number of the second sample recommendation models used for training the recommendation precision prediction model,
Figure BDA0003345560400000052
representing the proxy recommendation precision corresponding to the jth second sample recommendation model,
Figure BDA0003345560400000053
representing the agent recommendation precision, and alpha is a reference parameter;
the loss function satisfies the following condition:
Figure BDA0003345560400000054
wherein V represents a number of the second sample recommendation models used to train the recommendation accuracy prediction model,
Figure BDA0003345560400000061
representing the proxy recommendation precision corresponding to the jth second sample recommendation model,
Figure BDA0003345560400000062
representing the prediction recommendation precision corresponding to the jth second sample recommendation model at the tth moment, e(t)The target error corresponding to the T-th time is shown, and T shows T historical times.
In the embodiment of the disclosure, when the recommendation precision prediction model is trained, not only the first sample recommendation model is adopted to train the recommendation precision prediction model, but also the second sample recommendation model is adopted to train the recommendation precision prediction model, and in the training process, the proxy recommendation precision corresponding to the second sample recommendation model needs to meet the condition corresponding to the historical moment, that is, not only the training data in the current training process but also the training data at the historical moment are considered, and a self-evolution optimization framework aiming at the second sample recommendation model is added, so that the training efficiency is improved, and the accuracy of the recommendation precision prediction model obtained by training is also improved.
According to still another aspect of the embodiments of the present disclosure, there is provided a classification model processing method, including:
acquiring a classification model, wherein the classification model comprises a plurality of operators which are arranged in sequence, and the operators sequentially process data to be classified input into the classification model according to the sequence to obtain data types corresponding to the data;
based on the classification model, acquiring operator characteristics corresponding to each operator in the classification model and structural characteristics of the classification model, wherein the structural characteristics represent a connection relation between any two operators in the multiple operators;
performing feature extraction based on the operator features corresponding to each operator and the structural features to obtain model features corresponding to the classification models;
and performing precision prediction based on the model characteristics to obtain the classification precision corresponding to the classification model.
According to the method provided by the embodiment of the disclosure, the operator characteristics and the structural characteristics corresponding to the classification model are obtained firstly, and then the operator characteristics and the structural characteristics are utilized to extract effective model characteristics, so that the characteristics of the classification model can be fully obtained by the mode of obtaining the model characteristics, the obtained model characteristics can more accurately represent the classification model, the classification precision obtained based on the model characteristic prediction is more accurate, and the accuracy of the classification precision is improved.
According to still another aspect of the embodiments of the present disclosure, there is provided a model processing apparatus, the apparatus including:
the recommendation model obtaining unit is configured to execute obtaining of a recommendation model, the recommendation model comprises a plurality of operators which are arranged in sequence, the operators sequentially process account data and article data which are input into the recommendation model according to the sequence to obtain recommendation parameters, and the recommendation parameters indicate whether the articles are recommended to the accounts or not;
a first feature obtaining unit configured to perform obtaining, based on the recommendation model, an operator feature corresponding to each operator in the recommendation model and a structural feature of the recommendation model, where the structural feature represents a connection relationship between any two operators in the plurality of operators;
the second feature acquisition unit is configured to perform feature extraction based on the operator features corresponding to each operator and the structural features to obtain model features corresponding to the recommended model;
and the recommendation precision prediction unit is configured to perform precision prediction based on the model characteristics to obtain recommendation precision corresponding to the recommendation model.
In some embodiments, the recommendation model further includes an index identifier corresponding to each operator, and the first feature obtaining unit includes:
the operator characteristic obtaining subunit is configured to perform query on the basis of the index identifier corresponding to each operator to obtain the operator characteristic corresponding to each operator from an operator characteristic table;
the operator feature table comprises index identifications and operator features corresponding to at least one operator, wherein the operator features are obtained by performing feature extraction on the operators.
In some embodiments, the recommendation model includes N operators, where N is an integer greater than 1, and the first feature obtaining unit includes:
a structural feature obtaining subunit, configured to execute creating, based on the N operators in the recommendation model, a target matrix with N rows and N columns, where the N operators are respectively used as the N rows and N columns of the target matrix, rows and columns corresponding to elements on a diagonal line in the target matrix are the same operator, a value of an element on the diagonal line in the target matrix is equal to the number of other operators connected to the operator corresponding to the element, rows and columns corresponding to elements on other positions in the target matrix are different two operators, and a value of an element on the other position indicates whether the two operators corresponding to the element have a connection relationship;
the structural feature obtaining subunit is further configured to perform feature extraction on the target matrix to obtain the structural feature.
In some embodiments, the second feature obtaining unit includes:
the feature splicing subunit is configured to splice the operator features corresponding to each operator and the structural features to obtain splicing features corresponding to the recommendation model;
and the coding subunit is configured to perform coding on the splicing characteristics to obtain the model characteristics.
In some embodiments, the recommendation precision prediction model includes a feature extraction network and a prediction network,
the feature extraction network is used for performing feature extraction based on the operator features corresponding to each operator and the structural features to obtain the model features;
and the prediction network is used for predicting the accuracy of the model characteristics to obtain the recommendation accuracy.
In some embodiments, the recommendation accuracy prediction model further comprises a feature acquisition network,
the feature acquisition network is used for acquiring operator features corresponding to each operator in the recommendation model;
the feature acquisition network is further configured to acquire the structural features of the recommendation model.
In some embodiments, the recommendation model processing apparatus further includes:
the model training unit is configured to execute obtaining of a sample operator characteristic, a first sample structural characteristic and a first sample recommendation precision corresponding to each sample operator in a first sample recommendation model, wherein the first sample structural characteristic represents a connection relation between any two sample operators in the first sample recommendation model;
the model training unit is further configured to execute calling of the recommendation precision prediction model, and process the sample operator characteristics and the first sample structure characteristics corresponding to each sample operator in the first sample recommendation model to obtain first prediction recommendation precision;
the model training unit is further configured to perform training of the recommendation accuracy prediction model based on the first prediction recommendation accuracy and the first sample recommendation accuracy.
In some embodiments, the recommendation model processing apparatus further includes:
a recommendation unit configured to perform input of the account data and the item data to the recommendation model;
the recommending unit is further configured to execute calling of a first operator in the recommending model, process the account data and the item data, and obtain output data of the first operator;
and the recommending unit is further configured to execute the steps of taking the output data of the first operator as the input data of the next operator connected with the first operator, calling the next operator to process the output data until the output data of the last operator in the recommending model is obtained, and determining the output data of the last operator as the recommending parameter.
In some embodiments, the model obtaining unit is configured to perform obtaining a plurality of recommendation models, where operators in any two recommendation models are different, or connection relationships between operators in any two recommendation models are different;
the recommendation model processing apparatus further includes:
the model selecting unit is configured to execute the recommendation accuracy corresponding to the recommendation models, and select a target recommendation model corresponding to the maximum recommendation accuracy from the recommendation models;
and the recommending unit is configured to call the target recommending model for recommending.
In some embodiments, the recommendation model further includes a plurality of edges, each edge connects two operators, and the operators are configured to process input data input to the operators to obtain output data; and the output data of one operator connected by each edge is used as the input data of the other operator connected by each edge.
According to still another aspect of the embodiments of the present disclosure, there is provided a recommendation precision prediction model training apparatus, including:
the sample acquiring unit is configured to acquire a sample operator characteristic, a first sample structural characteristic and a first sample recommendation precision corresponding to each sample operator in a first sample recommendation model, wherein the first sample structural characteristic represents a connection relation between any two sample operators in the first sample recommendation model;
the precision prediction unit is configured to execute calling of a recommended precision prediction model, perform feature extraction on sample operator features and first sample structure features corresponding to each sample operator in the first sample recommended model to obtain model features corresponding to the first sample recommended model, and perform precision prediction based on the model features corresponding to the first sample recommended model to obtain first predicted recommended precision;
a model training unit configured to perform training of the recommendation accuracy prediction model based on the first prediction recommendation accuracy and the first sample recommendation accuracy.
In some embodiments, the sample acquiring unit is further configured to perform: obtaining sample operator characteristics, second sample structure characteristics and proxy recommendation precision corresponding to each sample operator in a second sample recommendation model, wherein the second sample structure characteristics represent a connection relation between any two sample operators in the second sample recommendation model, and the proxy recommendation precision is a variable for indicating the recommendation precision of the second sample recommendation model;
the precision prediction unit is configured to execute calling of the recommendation precision prediction model, perform feature extraction on the sample operator features and the second sample structure features corresponding to each sample operator in the second sample recommendation model to obtain model features corresponding to the second sample recommendation model, and perform precision prediction based on the model features corresponding to the second sample recommendation model to obtain second prediction recommendation precision;
the model training unit is configured to perform training of the recommendation accuracy prediction model based on the first prediction recommendation accuracy, the first sample recommendation accuracy, the second prediction recommendation accuracy, and the agent recommendation accuracy.
In some embodiments, the model training unit is configured to perform:
acquiring a first error between the first prediction recommendation precision and the first sample recommendation precision, and acquiring a second error between the second prediction recommendation precision and the second sample recommendation precision;
and adjusting model parameters in the recommendation precision prediction model and the proxy recommendation precision based on the first error and the second error so that the error between the prediction recommendation precision corresponding to the second sample recommendation model at the historical time and the adjusted proxy recommendation precision approaches to a target error corresponding to the historical time.
In some embodiments, the model training unit is configured to perform:
the recommendation accuracy prediction model is trained based on the first prediction recommendation accuracy, the first sample recommendation accuracy, the second prediction recommendation accuracy and the agent recommendation accuracy by adopting the following loss functions:
Figure BDA0003345560400000101
wherein L represents the loss function, fθ(. cndot.) represents the recommendation accuracy prediction model, θ represents a model parameter in the recommendation accuracy prediction model, fθ(xi) Representing the first prediction recommendation essence corresponding to the ith first sample recommendation modelDegree, xiRepresenting sample operator characteristics and first sample structure characteristics corresponding to each sample operator in the ith first sample recommendation model, n representing the number of the first sample recommendation models used for training the recommendation precision prediction model, yiRepresenting the first sample recommendation precision, f, corresponding to the ith first sample recommendation modelθ(vj) Represents the second prediction recommendation precision, v, corresponding to the jth second sample recommendation modeljRepresenting sample operator characteristics and the second sample structure characteristics corresponding to each sample operator in the jth second sample recommendation model, V representing the number of the second sample recommendation models used for training the recommendation precision prediction model,
Figure BDA0003345560400000102
representing the proxy recommendation precision corresponding to the jth second sample recommendation model,
Figure BDA0003345560400000103
representing the agent recommendation precision, and alpha is a reference parameter;
the loss function satisfies the following condition:
Figure BDA0003345560400000104
wherein V represents a number of the second sample recommendation models used to train the recommendation accuracy prediction model,
Figure BDA0003345560400000105
representing the proxy recommendation precision corresponding to the jth second sample recommendation model,
Figure BDA0003345560400000106
representing the prediction recommendation precision corresponding to the jth second sample recommendation model at the tth moment, e(t)The target error corresponding to the T-th time is shown, and T shows T historical times.
According to still another aspect of the embodiments of the present disclosure, there is provided a classification model processing apparatus including:
the classification model acquisition unit is configured to execute acquisition of a classification model, the classification model comprises a plurality of operators which are arranged in sequence, and the operators sequentially process data to be classified input into the classification model according to the sequence to obtain data types corresponding to the data;
a first feature obtaining unit configured to perform obtaining, based on the classification model, an operator feature corresponding to each operator in the classification model and a structural feature of the classification model, where the structural feature represents a connection relationship between any two operators in the plurality of operators;
the second feature acquisition unit is configured to perform feature extraction based on the operator features corresponding to each operator and the structural features to obtain model features corresponding to the classification models;
and the classification precision prediction unit is configured to execute precision prediction based on the model characteristics to obtain the classification precision corresponding to the classification model.
According to still another aspect of the embodiments of the present disclosure, there is provided an electronic apparatus including:
one or more processors;
a memory for storing the one or more processor-executable instructions;
wherein the one or more processors are configured to perform the recommendation model processing method of the above aspect, or configured to perform the recommendation accuracy prediction model training method of the above aspect, or configured to perform the classification model processing method of the above aspect.
According to still another aspect of the embodiments of the present disclosure, there is provided a computer-readable storage medium, wherein when executed by a processor of an electronic device, instructions of the storage medium enable the electronic device to execute the recommendation model processing method according to the above aspect, or to execute the recommendation accuracy prediction model training method according to the above aspect, or to execute the classification model processing method according to the above aspect.
According to yet another aspect of the embodiments of the present disclosure, there is provided a computer program product, which includes a computer program executed by a processor to implement the recommendation model processing method of the above aspect, or to implement the recommendation accuracy prediction model training method of the above aspect, or to implement the classification model processing method of the above aspect.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and together with the description, serve to explain the principles of the disclosure.
FIG. 1 is a flow diagram illustrating a recommendation model processing method in accordance with an exemplary embodiment.
FIG. 2 is a flow diagram illustrating another recommendation model processing method in accordance with an exemplary embodiment.
FIG. 3 is a flow diagram illustrating one approach to neural architecture searching, according to an example embodiment.
FIG. 4 is a schematic diagram illustrating a directed acyclic graph in accordance with an illustrative embodiment.
FIG. 5 is a schematic diagram illustrating an object matrix according to an exemplary embodiment.
FIG. 6 is a schematic diagram illustrating a recommendation accuracy prediction model in accordance with an exemplary embodiment.
FIG. 7 is a schematic diagram illustrating another recommendation accuracy prediction model in accordance with an exemplary embodiment.
FIG. 8 is a flowchart illustrating yet another recommendation model processing method according to an example embodiment.
FIG. 9 is a diagram illustrating a recommendation model processing method according to an example embodiment.
FIG. 10 is a schematic diagram illustrating a recommendation accuracy prediction model training method in accordance with an exemplary embodiment.
FIG. 11 is a diagram illustrating a recommendation accuracy variation according to an exemplary embodiment.
FIG. 12 is a flowchart illustrating a classification model processing method according to an exemplary embodiment.
Fig. 13 is a block diagram illustrating a recommendation model processing apparatus according to an example embodiment.
Fig. 14 is a block diagram illustrating another recommendation model processing apparatus according to an example embodiment.
FIG. 15 is a block diagram illustrating a recommendation accuracy prediction model training apparatus in accordance with an exemplary embodiment.
Fig. 16 is a block diagram illustrating a classification model processing apparatus according to an exemplary embodiment.
Fig. 17 is a block diagram illustrating a structure of a terminal according to an exemplary embodiment.
FIG. 18 is a block diagram illustrating a configuration of a server according to an example embodiment.
Detailed Description
In order to make the technical solutions of the present disclosure better understood by those of ordinary skill in the art, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.
It should be noted that the terms "first," "second," and the like in the description and claims of the present disclosure and in the description of the above-described figures are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the disclosure described herein are capable of operation in sequences other than those illustrated or otherwise described herein. The implementations described in the exemplary embodiments below are not intended to represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.
It should be noted that, as used in this disclosure, the terms "at least one," "a plurality," "each," "any," and the like, at least one includes one, two, or more than two, and a plurality includes two or more than two, each referring to each of the corresponding plurality, and any referring to any one of the plurality. For example, the plurality of operators includes 3 operators, and each operator refers to each of the 3 operators, and any one refers to any one of the 3 operators, which may be a first operator, a second operator, or a third operator.
It should be noted that, the data (including but not limited to user device information, user personal information, etc.) referred to in the present disclosure is information authorized by the user or sufficiently authorized by each party.
The execution subject of the present application is an electronic device. Optionally, the electronic device is a terminal or a server. The terminal can be various terminals such as a mobile phone, a tablet computer, a computer and the like, and the server is a server, or a server cluster consisting of a plurality of servers, or a cloud computing service center.
Fig. 1 is a flowchart illustrating a recommendation model processing method according to an exemplary embodiment, and referring to fig. 1, an execution subject of the method is an electronic device, and the method includes the following steps:
in step 101, the electronic device obtains a recommendation model.
The recommendation model comprises a plurality of operators which are arranged in sequence, the operators process the account data and the article data which are input into the recommendation model in sequence to obtain recommendation parameters, and the recommendation parameters indicate whether the articles are recommended to the account. Each operator is used for processing input data input to the operator to obtain output data, the input data of the first operator are account data and article data, and the input data of other operators except the first operator are the output data of the last operator connected with the other operators.
In step 102, the electronic device obtains, based on the recommendation model, an operator characteristic corresponding to each operator in the recommendation model and a structural characteristic of the recommendation model.
The operator features are used for describing corresponding operators, and the structural features of the recommendation model represent the connection relation between any two operators in the operators, namely represent the topological structure information of the recommendation model.
In step 103, the electronic device performs feature extraction based on the operator features and the structural features corresponding to each operator to obtain model features corresponding to the recommended model.
For a recommendation model, in order to obtain accurate recommendation accuracy, first of all, a model feature corresponding to the recommendation model needs to be extracted, in the embodiment of the present disclosure, an operator feature and a structural feature corresponding to each operator are obtained first, and the operator feature and the structural feature corresponding to each operator represent part of features of the recommendation model, so that in order to obtain accurate features for describing the recommendation model, further feature extraction needs to be performed on the operator feature and the structural feature corresponding to each operator, so that a plurality of operator features and structural features are fully fused together to obtain the model feature corresponding to the recommendation model.
In step 104, the electronic device performs precision prediction based on the model features to obtain recommendation precision corresponding to the recommendation model.
The electronic equipment processes the model features so as to predict recommendation precision corresponding to the recommendation model, wherein the recommendation precision is used for expressing the accuracy of the recommendation model. Thereafter, it may be determined whether the recommendation model is actually applicable based on the recommendation accuracy.
According to the method provided by the embodiment of the disclosure, the operator characteristics and the structural characteristics corresponding to the recommendation model are obtained firstly, and then the effective model characteristics are extracted by utilizing the operator characteristics and the structural characteristics, so that the characteristics of the recommendation model can be fully obtained by the mode of obtaining the model characteristics, the obtained model characteristics can more accurately represent the recommendation model, the recommendation precision obtained based on the model characteristics is more accurate, and the accuracy of the recommendation precision is improved.
Fig. 2 is a flowchart illustrating another recommendation model processing method according to an exemplary embodiment, and referring to fig. 2, the method is executed by an electronic device, and includes the following steps:
in step 201, the electronic device obtains a recommendation model.
The recommendation model comprises a plurality of operators and a plurality of edges which are arranged in sequence, wherein each edge is connected with two operators, each operator is connected with at least one edge, namely each operator is connected with other operators through at least one edge, namely one operator can be connected with the other operator through one edge, and the operators can also be respectively connected with the operators through a plurality of edges. The operators are used for processing input data input to the operators to obtain output data, and the output data of one operator connected with each edge is used as the input data of the other operator connected with each edge. The operator may be a convolution operator, an deconvolution operator, an activation function, a normalization operator, or other operators.
For the recommendation model, input data of the recommendation model are account data and item data, wherein the account data includes account information, description information corresponding to the account, or other data related to the account, and the item data includes an item name, item description information, or other data related to the account. Taking recommending commodities to users as an example, the account data comprises an account number and a user label corresponding to the user, and the article data comprises information such as commodity names, commodity description information and commodity prices. The output data of the recommendation model is a recommendation parameter indicating whether to recommend the item to the account, for example, if the recommendation parameter is 1, it indicates that the item is recommended to the account, and if the recommendation parameter is 0, it indicates that the item is not recommended to the account.
And the plurality of operators sequentially process the account data and the article data input into the recommendation model according to the sequence to obtain recommendation parameters. In some embodiments, the electronic device inputs account data and item data into the recommendation model; calling a first operator in the recommendation model, and processing the account data and the article data to obtain output data of the first operator; and taking the output data of the first operator as the input data of the next operator connected with the first operator, calling the next operator to process the output data until the output data of the last operator in the recommendation model is obtained, and determining the output data of the last operator as the recommendation parameter. The input data of the first operator in the recommendation model is account data and article data, and the input data of other operators except the first operator is output data of the last operator connected with the other operators. The operators are sequentially processed, so that each operator can fully play a role, and the characteristics of the account data and the article data are gradually extracted in the processing process, so that accurate recommendation parameters are obtained, and article recommendation based on the recommendation model is realized.
In some embodiments, the recommendation model further comprises an index identifier corresponding to each operator, the index identifier being used to uniquely represent the corresponding one of the operators. For example, the index is identified as an operator name, a label, and the like.
In some embodiments, the electronic device obtains the recommendation model by way of a neural architecture search. Among them, Neural Architecture Search (NAS) refers to automatically finding out a recommendation model with excellent effect in a predefined Search space. Referring to the neural architecture search process shown in fig. 3, the neural architecture search includes three parts, namely a search space, a search strategy and a model effect evaluation, where the search space is a predefined space including a plurality of operators, the search strategy is predefined and is used to indicate how to obtain a recommendation model from one operator in the search space, a corresponding search strategy can be defined according to the recommendation model to be searched, and the model effect evaluation refers to predicting recommendation accuracy corresponding to the recommendation model obtained by the search so as to determine recommendation accuracy corresponding to the recommendation model obtained by the search. The disclosed embodiments do not limit the process of building the recommendation model.
In some embodiments, the recommendation model is a manually designed recommendation model.
In addition, the recommendation model may also be sent to the electronic device by other electronic devices, and the embodiment of the present disclosure does not limit the obtaining manner of the recommendation model.
In some embodiments, the recommendation model is regarded as a Directed Acyclic Graph (DAG), a node in the DAG is an operator in the recommendation model, an edge in the DAG is a connection relationship between two operators, that is, an edge between two nodes indicates that two operators corresponding to the two nodes are connected, and an edge between two nodes indicates that two operators corresponding to the two nodes are not connected. For example, see the directed acyclic graph 1 corresponding to the recommendation model 1 and the directed acyclic graph 2 corresponding to the recommendation model 2 shown in fig. 4, where the directed acyclic graph 1 includes an input operator, an output operator, four convolution operators of 3 × 3, and two convolution operators of 1 × 1, the operators are connected in the order shown in fig. 4, an edge between the two operators is represented by an arrow, a direction indicated by the arrow is a data flow direction, that is, output data of an operator connected to a tail of the arrow is input data of an operator pointed by a head of the arrow, and the directed acyclic graph 2 includes the input operator, the output operator, four convolution operators of 3 × 3, and an MP (Matching Pursuits) operator. And then, when the structural features of the recommendation model are obtained, the structural features can be obtained based on the directed acyclic graph corresponding to the recommendation model.
It should be noted that, the number and type of operators included in the recommendation model and the connection relationship between the operators are not limited in the embodiments of the present disclosure, for example, referring to fig. 4, the number of operators, the type of operators, and the connection relationship between the operators in the two recommendation models shown in fig. 4 are not the same.
In step 202, the electronic device queries, based on the index identifier corresponding to each operator in the recommendation model, an operator feature corresponding to each operator from the operator feature table.
The operator feature table comprises index identifications and operator features corresponding to at least one operator, namely, a corresponding relation exists between the index identifications and the operator features, and the electronic equipment can inquire the operator features corresponding to the index identifications from the operator feature table based on the index identifications. The operator features are used for describing corresponding operators, and the operator features are obtained by performing feature extraction on the operators, and the operator features are vectors, matrixes or other forms, for example, the operator features are matrixes with N rows and M columns. The embodiments of the present disclosure do not limit the form of the operator features.
Because the operator features in the operator feature table are extracted in advance, the operator features can be quickly obtained by inquiring the operator feature table without temporarily extracting the operator features corresponding to the operators, and therefore the efficiency of obtaining recommendation precision is improved.
In some embodiments, the operator feature table is trainable, that is, after the electronic device obtains the operator feature table, the operator feature in the operator feature table may be continuously updated, so that the accuracy of each operator feature in the operator feature table is continuously improved, and the operator feature can accurately describe the corresponding operator.
It should be noted that, the embodiment of the present disclosure is only described by taking an example of querying an operator feature from an operator feature table, and in another embodiment, the electronic device may obtain the operator feature corresponding to each operator in other manners. For example, the operator features corresponding to the operators are extracted by using the feature extraction model.
In step 203, the electronic device creates an object matrix based on a plurality of operators in the recommendation model, and performs feature extraction on the object matrix to obtain the structural features of the recommendation model.
Taking the recommendation model including N operators as an example, where N is an integer greater than 1, and accordingly, the electronic device creates a target matrix with N rows and N columns based on the N operators in the recommendation model. The N operators are respectively used as N rows and N columns of the target matrix, the rows and the columns corresponding to the elements on the diagonal line in the target matrix are the same operators, the value of the element on the diagonal line in the target matrix is equal to the number of other operators connected with the operator corresponding to the element, the rows and the columns corresponding to the elements on other positions in the target matrix are two different operators, and the value taking of the elements on other positions indicates whether the two operators corresponding to the elements have a connection relation.
For example, referring to the target matrix shown in fig. 5, the recommendation model includes N operators, a first row in the target matrix corresponds to a first operator, a first column corresponds to the first operator, an element in the first row and the first column in the target matrix is the number of operators connected to the first operator, an element in the first row and the second column in the target matrix is-1, which indicates that the first operator is connected to the second operator, and an element in the first row and the nth column is 0, which indicates that the first operator is not connected to the nth operator.
In some embodiments, since there is a direction between two operators connected in the recommendation model, elements at other locations in the objective matrix can also represent the direction of the connection between the two operators. For example, in the target matrix, a first row and a first column correspond to operator 1, a second row and a second column correspond to operator 2, the element corresponding to the first row and the second column is-1, that is, operator 1 is connected with operator 2, and the connection direction is from operator 1 to operator 2, that is, the output of operator 1 is the input of operator 2; the element in the second row and the first column is 0, which means that operator 1 is not connected to operator 2, or the connection direction of operator 1 to operator 2 is not from operator 2 to operator 1.
In some embodiments, when a recommendation model is regarded as a directed acyclic graph, a corresponding degree matrix and an adjacency matrix are obtained based on the recommendation model, a value of an element on a diagonal line in the degree matrix is equal to the number of operators connected to one operator corresponding to the element, a value of an element in the adjacency matrix indicates whether two operators corresponding to the element have a connection relationship, and a difference value between the degree matrix and the adjacency matrix is used as a target matrix.
In the embodiment of the present disclosure, the target matrix only represents a connection relationship between any two operators in the multiple operators, that is, the target matrix represents a relative position relationship before each operator, and for convenience of subsequent processing, the target matrix is subjected to feature extraction to obtain a structural feature. The structural features are distinguished from the object matrix in that the structural features are continuous coding features, whereas the object matrix is only discrete data. Because the target matrix can fully represent the connection relation between any two operators in the recommendation model, namely the topological structure information of the recommendation model is fully obtained, the structural features obtained based on the target matrix can accurately represent the structure of the recommendation model, and the accuracy of the structural features is improved.
It should be noted that, the embodiment of the present disclosure is only described as performing step 202 first and then step 203, in another embodiment, step 203 may be performed first and then step 202 is performed, or step 202 and step 203 may be performed simultaneously, and the order of performing step 202 and step 203 is not limited in the embodiment of the present disclosure.
In step 204, the electronic device splices the operator features and the structural features corresponding to each operator to obtain the splice features corresponding to the recommended model.
In the embodiment of the present disclosure, in order to ensure that the operator features and the structural features can be spliced, in step 203, when the feature of the target matrix is extracted, the dimension of the obtained structural features is the same as the dimension of the operator features, so that the structural features and the operator features with the same dimension can be spliced.
In some embodiments, the electronic device directly adds the structural feature to the plurality of operator features to obtain a mosaic feature, or adds the structural feature to the plurality of operator features to obtain the mosaic feature; or the multiple operator characteristics and structural characteristics can be spliced in other modes.
In the embodiment of the disclosure, the plurality of operator features and the structural features are respectively obtained, the obtained operator features can accurately represent operators in the recommendation model, and the structural features can accurately represent the structure of the recommendation model, so that the model features obtained by splicing and encoding the plurality of operator features and the structural features can accurately represent the recommendation model, and the accuracy of the model features is improved.
In step 205, the electronic device encodes the splicing feature to obtain a model feature corresponding to the recommended model.
In the embodiment of the present disclosure, although the splicing features include features and structural features of each operator of the recommendation model, the splicing features cannot directly represent an effect generated by connection or disconnection of each operator in the recommendation model. Therefore, in order to obtain accurate characteristics representing the recommended model and encode the splicing characteristics, the operator characteristics corresponding to each operator and the structural characteristics of the recommended model are fused in the encoding process to obtain the model characteristics. The model features obtained by the method can accurately represent the recommended model, and the accuracy of the model features is improved.
In some embodiments, the stitching features are processed based on an attention mechanism, a weight of each operator is determined, and the operator features corresponding to the operators are adjusted based on the weights of the operators. For example, based on the weight of the operator, the features corresponding to the operator are weighted to obtain the weighted operator features. Alternatively, the attention mechanism is a multi-head attention mechanism or other type of attention mechanism.
In step 206, the electronic device performs accuracy prediction based on the model features to obtain recommendation accuracy corresponding to the recommendation model.
And the electronic equipment carries out precision prediction on the recommendation model based on the model characteristics to obtain the recommendation precision of the recommendation model, wherein the recommendation precision represents the accuracy of the recommendation model. The higher the recommendation precision is, the more accurate the recommendation model is, and the recommendation model can be practically applied; the smaller the recommendation precision is, the more inaccurate the recommendation model is, and the recommendation model cannot be practically applied.
In some embodiments, in a scenario of neural architecture search, after the electronic device predicts recommendation accuracy corresponding to a recommendation model, it may be determined whether the recommendation model is actually applicable according to the recommendation accuracy, and if the recommendation model is not actually applicable, it is necessary to continue searching for other recommendation models.
It should be noted that the recommendation model in the embodiment of the present disclosure may be a trained recommendation model or an untrained recommendation model, which is not limited in the embodiment of the present disclosure.
It should be noted that, in another embodiment, the above embodiment is only described by taking an example of processing one recommendation model, and in another embodiment, the electronic device obtains a plurality of recommendation models, where any two recommendation models in the plurality of recommendation models have different operators, or connection relationships between operators in any two recommendation models are different, where different connection relationships between operators means that connection relationships between any two operators in the plurality of operators are different. For example, there are five recommendation models, in which there are different operators in the first and second recommendation models, the same operator in the third and fourth recommendation models, but the connection relationships between the operators are different, and there are different operators in the fourth and fifth recommendation models.
Then, for each recommendation model, the recommendation model processing method shown in the above steps 202 to 206 is respectively adopted to obtain the recommendation precision corresponding to each recommendation model, and then, based on the recommendation precision corresponding to the multiple recommendation models, a target recommendation model corresponding to the maximum recommendation precision is selected from the multiple recommendation models, and the target recommendation model is called to perform recommendation, that is, the target recommendation model is applied to an actual article recommendation scene. In the embodiment of the disclosure, based on the recommendation precision corresponding to the plurality of recommendation models, the most accurate recommendation model can be selected from the plurality of recommendation models, so that when the selected recommendation model is applied to an actual recommendation scene, accurate recommendation can be performed based on the recommendation model, and the recommendation accuracy is improved.
According to the method provided by the embodiment of the disclosure, the operator characteristics and the structural characteristics corresponding to the recommendation model are obtained firstly, and then the effective model characteristics are extracted by utilizing the operator characteristics and the structural characteristics, so that the characteristics of the recommendation model can be fully obtained by the mode of obtaining the model characteristics, the obtained model characteristics can more accurately represent the recommendation model, the recommendation precision obtained based on the model characteristics is more accurate, and the accuracy of the recommendation precision is improved.
In addition, since the operator features in the operator feature table are extracted in advance, the operator features can be quickly acquired by inquiring the operator feature table, so that the efficiency of acquiring recommendation precision is improved. The target matrix can fully represent the connection relation between any two operators in the recommendation model, namely the topological structure information of the recommendation model is fully obtained, so that the structure of the recommendation model can be accurately represented by the structural features obtained based on the target matrix, and the accuracy of the structural features is improved.
In addition, in the related art, a large amount of computing resources are required for traditional NAS methods such as reinforcement learning and evolutionary learning to train each recommendation model to determine recommendation accuracy corresponding to each recommendation model. Recently, the search time of the NAS method based on micro optimization is greatly reduced compared with the conventional search method, but due to the discontinuity of the optimization space, the optimization performance is usually collapsed, so that the optimal model which meets the conditions cannot be searched out. Compared with the method that the recommendation model obtained by searching is trained and then the recommendation precision is tested by adopting test data, the method provided by the embodiment of the disclosure can directly predict the precision of the recommendation model obtained by searching without training the recommendation model, thereby improving the efficiency of obtaining the recommendation precision, saving the time of model training and avoiding the model which cannot be practically applied by time waste training.
In another embodiment, the electronic device can call the recommendation precision prediction model to process the recommendation model, so as to obtain the recommendation precision corresponding to the recommendation model.
Referring to fig. 6, the recommendation accuracy prediction model includes a feature extraction network 601 and a prediction network 602. The feature extraction network 601 is configured to extract model features corresponding to recommendation models, and the prediction network 602 is configured to predict recommendation accuracy corresponding to recommendation models.
In some embodiments, referring to fig. 7, the recommendation accuracy prediction model further includes a feature obtaining network 603, where the feature obtaining network 603 is configured to obtain an operator feature corresponding to each operator in the recommendation model and a structural feature of the recommendation model.
Fig. 8 is a flowchart illustrating another recommendation model processing method according to an exemplary embodiment, and referring to fig. 8, the method is executed by an electronic device, and includes the following steps:
in step 801, the electronic device obtains a recommendation model.
The implementation of step 801 is the same as the implementation of step 201, and is not described herein again.
In step 802, the electronic device invokes a feature acquisition network to acquire an operator feature corresponding to each operator in the recommendation model.
In some embodiments, the index identifier corresponding to each operator in the recommendation model is used as an input of the recommendation precision prediction model, the electronic device invokes the feature acquisition network, and based on the index identifier corresponding to each operator, the operator feature corresponding to each operator is queried from the operator feature table.
In step 803, the electronic device invokes the feature acquisition network to acquire the structural features of the recommendation model.
In some embodiments, the target matrix corresponding to the recommendation model is used as an input of the recommendation precision prediction model, the electronic device calls the feature acquisition network, and feature extraction is performed on the target matrix to obtain the structural features. The feature acquisition network can convert a discrete target matrix into a continuous structural feature, and optionally, the feature acquisition network is an MLP (Multi-Layer Perceptron), a convolutional neural network, or another type of neural network, which is not limited in this disclosure.
In step 804, the electronic device invokes a feature extraction network, and performs feature extraction based on the operator features and the structural features corresponding to each operator to obtain model features corresponding to the recommended model.
In some embodiments, the feature extraction network includes a splicing layer and a feature extraction layer, the electronic device calls the splicing layer to splice operator features and structural features corresponding to each operator to obtain splicing features corresponding to the recommended model, and calls the feature extraction layer to encode the splicing features to obtain model features.
Optionally, the feature extraction layer is a transformer encoder, and the feature extraction layer includes an attention unit and an encoding unit, and calls the attention unit to process a plurality of operator features to obtain weights of the operators, processes the operator features corresponding to the operators based on the weights of the operators to obtain weighted operator features corresponding to each operator, and then calls the encoding unit to encode the weighted operator features and structural features to obtain model features. Wherein the attention unit is a multi-head attention unit or other types of attention units.
In another embodiment, the feature extraction network may be other types of encoders.
In some embodiments, the recommendation precision prediction model includes a plurality of feature extraction networks, for example, 4 feature extraction networks. The electronic device takes the output of the feature acquisition network as the input of a first feature extraction network, then takes the output of the first feature extraction network as the input of a next feature extraction network until the output of the last feature extraction network is obtained, and takes the output as the model feature.
In step 805, the electronic device invokes a prediction network to perform accuracy prediction on the model features to obtain recommendation accuracy corresponding to the recommendation model.
Wherein the prediction network is an MLP or other neural network.
In some embodiments, the recommendation accuracy prediction model includes a plurality of prediction networks, for example, 2 prediction networks. The electronic equipment takes the output of the characteristic acquisition network as the input of a first prediction network, then takes the output of the first prediction network as the input of a next prediction network until the output of the last prediction network is obtained, and takes the output as the prediction precision.
Optionally, the output of the recommendation accuracy prediction model is a score, and the score is used to represent the recommendation accuracy.
See, for example, the schematic diagram of the recommendation model process shown in FIG. 9. Taking an index identification corresponding to each operator in a recommendation model and a target matrix corresponding to the recommendation model as input of a recommendation precision prediction model, then obtaining operator characteristics corresponding to each operator by inquiring an operator characteristic table based on the index identification corresponding to each operator, performing characteristic extraction on the target matrix through MLP to obtain structural characteristics, splicing the operator characteristics corresponding to each operator and the structural characteristics, inputting spliced characteristics obtained by splicing to a transformer encoder to obtain model characteristics, and then inputting the model characteristics to the MLP to obtain final recommendation precision.
In the embodiment of the disclosure, the recommendation precision is predicted by using the recommendation precision prediction model, the operator features and the structural features corresponding to the recommendation model are obtained first, and then the effective model features are extracted by using the operator features and the structural features.
Compared with the mode that the recommendation model is trained in the related art, the recommendation model is tested by adopting the test data, and the recommendation precision is determined according to the test result, the recommendation precision is predicted by utilizing the recommendation precision prediction model, so that the recommendation precision can be quickly obtained, and the efficiency of obtaining the recommendation precision is improved.
Before the recommendation accuracy prediction model is used to predict the recommendation accuracy, the recommendation accuracy prediction model needs to be trained, and a training process of the recommendation accuracy prediction model is described below by taking a training process as an example.
Fig. 10 is a flowchart illustrating another model training method according to an exemplary embodiment, and referring to fig. 10, the method is executed by an electronic device, and includes the following steps:
in step 1001, the electronic device obtains a sample operator feature, a first sample structural feature, and a first sample recommendation precision corresponding to each sample operator in the first sample recommendation model.
The first sample structural feature represents a connection relation between any two sample operators in the first sample recommendation model.
In step 1002, the electronic device calls a recommendation precision prediction model, and processes a sample operator characteristic corresponding to each sample operator in the first sample recommendation model and the first sample structure characteristic to obtain a first prediction recommendation precision.
In some embodiments, the electronic device calls a recommendation precision prediction model, performs feature extraction on sample operator features and first sample structure features corresponding to each sample operator in the first sample recommendation model to obtain model features corresponding to the first sample recommendation model, and performs precision prediction based on model features corresponding to the second sample recommendation model to obtain first prediction recommendation precision.
In step 1003, the electronic device obtains a sample operator characteristic, a second sample structure characteristic, and an agent recommendation precision corresponding to each sample operator in the second sample recommendation model.
The second sample structure feature represents a connection relation between any two sample operators in the second sample recommendation model, and the proxy recommendation precision is a variable used for indicating recommendation precision corresponding to the second sample recommendation model, namely the proxy recommendation precision can be changed in the training process.
In step 1004, the electronic device invokes a recommendation precision prediction model, and processes the sample operator characteristics and the second sample structural characteristics corresponding to each sample operator in the second sample recommendation model to obtain a second prediction recommendation precision.
In some embodiments, the electronic device calls a recommendation precision prediction model, performs feature extraction on the sample operator features and the second sample structure features corresponding to each sample operator in the second sample recommendation model to obtain model features corresponding to the second sample recommendation model, and performs precision prediction based on the model features corresponding to the second sample recommendation model to obtain second prediction recommendation precision.
In step 1005, the electronic device obtains a first error between the first predicted recommendation accuracy and the first sample recommendation accuracy, and a second error between the second predicted recommendation accuracy and the second sample recommendation accuracy.
In step 1006, the electronic device adjusts the model parameter and the proxy recommendation precision in the recommendation precision prediction model based on the first error and the second error, so that an error between the prediction recommendation precision corresponding to the historical time and the adjusted proxy recommendation precision of the second sample recommendation model approaches to a target error corresponding to the historical time.
In the embodiment of the disclosure, the electronic device trains the recommendation accuracy prediction model based on the first prediction recommendation accuracy, the first sample recommendation accuracy, the second prediction recommendation accuracy and the agent recommendation accuracy. In the training process, the first error is required to approach 0, the second error approaches 0, and the proxy recommendation precision meets the condition that the error between the prediction recommendation precision corresponding to the second sample recommendation model at the historical time and the adjusted proxy recommendation precision approaches the target error corresponding to the historical time. The target error is an error between the predicted recommendation accuracy of the second sample recommendation model at the historical time and the recommendation accuracy of the second sample recommendation model at the historical time. The historical time is the time when the second sample recommendation model is adopted to train the recommendation accuracy prediction model.
It should be noted that, in the above training process, only the operator features and the structural features are taken as input of the recommendation precision prediction model, in another embodiment, when the index identifier corresponding to the operator and the target matrix corresponding to the recommendation model are taken as input of the recommendation precision prediction model, the electronic device obtains the sample index identifier, the first sample matrix and the first sample recommendation precision corresponding to each sample operator in the first sample recommendation model, calls the recommendation precision prediction model, and processes the sample index identifier and the first sample matrix corresponding to each sample operator in the first sample recommendation model to obtain the first prediction recommendation precision; the electronic equipment obtains the sample index identification, the second sample matrix and the proxy recommendation precision corresponding to each sample operator in the second sample recommendation model, calls the recommendation precision prediction model, and processes the sample index identification and the second sample matrix corresponding to each sample operator in the second sample recommendation model to obtain the second prediction recommendation precision. The subsequent processing is the same as in steps 1005 and 1006 described above.
It should be noted that, in the above embodiment, the recommendation accuracy prediction model is trained by using the first sample recommendation model and the second sample recommendation model, but in another embodiment, the electronic device can train the recommendation accuracy prediction model by using the first sample recommendation model.
In addition, in some embodiments, the first sample recommendation model is multiple and the second sample recommendation model is multiple. The electronic equipment trains a recommendation precision prediction model based on the first prediction recommendation precision, the first sample recommendation precision, the second prediction recommendation precision and the agent recommendation precision by adopting the following loss functions:
Figure BDA0003345560400000231
wherein L represents a loss function, fθ(. cndot.) represents a recommended accuracy prediction model, θ represents a model parameter in the recommended accuracy prediction model, fθ(xi) Represents the first prediction recommendation precision, x, corresponding to the ith first sample recommendation modeliRepresenting sample operator characteristics and first sample structure characteristics corresponding to each sample operator in the ith first sample recommendation model, n representing the number of first sample recommendation models used for training recommendation precision prediction models, yiRepresenting a first sample recommendation precision, f, corresponding to the ith first sample recommendation modelθ(vj) Represents a second prediction recommendation precision, v, corresponding to the jth second sample recommendation modeljRepresenting the sample operator characteristic and the second sample structure characteristic corresponding to each sample operator in the jth second sample recommendation model, V representing the number of second sample recommendation models used for training the recommendation precision prediction model,
Figure BDA0003345560400000232
representing the proxy recommendation precision corresponding to the jth second sample recommendation model,
Figure BDA0003345560400000233
representing the proxy recommendation precision, alpha is a reference parameter, such as alpha is 0.1, 0.2, 0.3 or other numerical value;
the loss function satisfies the following condition:
Figure BDA0003345560400000234
wherein V represents the number of second sample recommendation models used to train the recommendation accuracy prediction model,
Figure BDA0003345560400000235
representing the proxy recommendation precision corresponding to the jth second sample recommendation model,
Figure BDA0003345560400000236
representing the prediction recommendation precision corresponding to the jth second sample recommendation model at the tth moment, e(t)The target error corresponding to the T-th time is shown, and T shows T historical times.
Optionally, the lagrange multiplier method is adopted to transform the above two formulas into the following formula, thereby converting the above formula into a minmax optimization problem:
Figure BDA0003345560400000237
where λ is the lagrange multiplier, and the other definitions are the same as above.
Then, the gradient descending and gradient ascending modes are adopted to solve the formula
Figure BDA0003345560400000238
The above formula is derived to obtain:
Figure BDA0003345560400000239
Figure BDA00033455604000002310
Figure BDA0003345560400000241
wherein k represents a k-th order partial derivative, ηθ、
Figure BDA0003345560400000242
and ηλFor the purpose of reference to the parameters,
Figure BDA0003345560400000243
represent to
Figure BDA0003345560400000244
The k-order partial derivative of theta is calculated,
Figure BDA0003345560400000245
represent to
Figure BDA0003345560400000246
To find
Figure BDA0003345560400000247
The k-th order partial derivative of (c),
Figure BDA0003345560400000248
represent to
Figure BDA0003345560400000249
And solving k-order partial derivatives of lambda.
See, for example, the schematic of the training process shown in fig. 11. In the process of training in the above manner, the recommendation precision corresponding to the second sample recommendation model is unchanged, the prediction recommendation precision of the second sample recommendation model at a plurality of historical moments is continuously changed and gradually approaches the recommendation precision, and the agent recommendation precision is also continuously changed and gradually approaches the recommendation precision.
According to the method provided by the embodiment of the disclosure, the recommendation precision prediction model for predicting the recommendation precision is trained, then, the recommendation precision prediction model can be utilized to process each operator characteristic and structural characteristic corresponding to the recommendation model, and effective model characteristics are extracted by utilizing the operator characteristics and the structural characteristics, so that the characteristics of the recommendation model are fully obtained, the obtained model characteristics can more accurately represent the recommendation model, the recommendation precision obtained by the recommendation precision prediction model based on the model characteristics is more accurate, and the accuracy of the recommendation precision is improved.
In the embodiment of the disclosure, when the recommendation precision prediction model is trained, not only the first sample recommendation model is adopted to train the recommendation precision prediction model, but also the second sample recommendation model is adopted to train the recommendation precision prediction model, so that more sample models for training are provided, and the accuracy and generalization capability of the recommendation precision prediction model obtained by training are improved.
When the recommendation precision prediction model is trained, the first sample recommendation model is adopted to train the recommendation precision prediction model, the second sample recommendation model is also adopted to train the recommendation precision prediction model, and in the training process, the agent recommendation precision corresponding to the second sample recommendation model needs to meet the conditions corresponding to the historical time, namely the training data in the current training process and the training data at the historical time are considered, and a self-evolution optimization framework aiming at the second sample recommendation model is added, so that the training efficiency is improved, and the accuracy of the recommendation precision prediction model obtained by training is also improved. In addition, the self-evolution optimization framework is adopted, and the recommendation precision prediction model is trained by adopting more sample recommendation models, so that the generalization capability of the recommendation precision prediction model is improved.
Fig. 12 is a flowchart illustrating a classification model processing method according to an exemplary embodiment, and referring to fig. 12, the method is executed by an electronic device, and includes the following steps:
in step 1201, the electronic device obtains a classification model.
The classification model comprises a plurality of operators which are arranged in sequence, and the operators sequentially process the data to be classified input into the classification model according to the sequence to obtain the data types corresponding to the data. For example, the classification model is used to classify an image to determine whether the image belongs to a face image, and the data type output by the classification model is a face image or a non-face image. Or, the classification model is used for classifying the article to determine the article class to which the article belongs; or the classification model is used to classify the audio to determine the type of audio to which the audio belongs. In the embodiment of the present disclosure, the role of the classification model and the specific structure of the classification model are not limited.
In some embodiments, the electronic device inputs data to be classified into a recommendation model; calling a first operator in the classification model, and processing data to be classified to obtain output data of the first operator; and taking the output data of the first operator as the input data of the next operator connected with the first operator, calling the next operator to process the output data until the output data of the last operator in the classification model is obtained, and determining the output data of the last operator as the recommended parameter. The input data of a first operator in the classification model is data to be classified, and the input data of other operators except the first operator is the output data of a last operator connected with the other operators.
In addition, the classification model is obtained in the same manner as the recommended model in the above embodiment, for example, the classification model is obtained by adopting a neural architecture search method, or the classification model is a model designed manually.
In step 1202, the electronic device obtains, based on the classification model, an operator feature corresponding to each operator in the classification model and a structural feature of the classification model.
Wherein the structural feature represents a connection relationship between any two operators of the plurality of operators.
In step 1203, the electronic device performs feature extraction based on the operator features and the structural features corresponding to each operator, so as to obtain model features corresponding to the classification model.
The embodiment of obtaining the model features corresponding to the classification model in steps 1202 to 1203 is the same as the embodiment of obtaining the model features corresponding to the recommendation model in steps 202 to 205 of the above embodiment. In contrast, the classification model is different from operators or connection relations in the recommendation model.
In step 1204, the electronic device performs accuracy prediction based on the model features to obtain classification accuracy corresponding to the classification model.
Wherein, the classification precision is used for representing the classification accuracy of the classification model. The higher the classification precision is, the more accurate the classification model is, and the classification model can be practically applied; the smaller the classification precision is, the more inaccurate the classification model is, and the classification model cannot be practically applied.
It should be noted that, the foregoing embodiment is only described by taking an example of processing one classification model, and in another embodiment, the electronic device obtains a plurality of classification models, where a plurality of operators in any two classification models in the plurality of classification models and a connection relationship between any two operators in the plurality of operators are different. Then, for each classification model, the classification accuracy corresponding to each classification model is obtained by respectively adopting the classification model processing modes shown in the above steps 1202 to 1204, and then, based on the classification accuracy corresponding to the plurality of classification models, a target classification model corresponding to the maximum classification accuracy is selected from the plurality of classification models, and the target classification model is called for classification.
According to the method provided by the embodiment of the disclosure, the operator characteristics and the structural characteristics corresponding to the classification model are obtained firstly, and then the operator characteristics and the structural characteristics are utilized to extract effective model characteristics, so that the characteristics of the classification model can be fully obtained by the mode of obtaining the model characteristics, the obtained model characteristics can more accurately represent the classification model, the classification precision obtained based on the model characteristic prediction is more accurate, and the accuracy of the classification precision is improved.
In addition, in another embodiment, the electronic device can call the classification accuracy prediction model to process the plurality of operator features and the structural features corresponding to the classification model, so as to obtain the classification accuracy corresponding to the classification model. The structure of the classification precision prediction model is the same as that of the recommendation precision prediction model in the above embodiment, except that the classification precision prediction model is used for processing the model features corresponding to the classification model to obtain the classification precision. The training process of the classification precision prediction model is the same as the training process of the recommended precision prediction model in the embodiment, but the classification precision prediction model is trained by using a sample classification model, and the capability of the classification precision prediction model capable of predicting the classification precision corresponding to the classification model is trained.
In another embodiment, a processing procedure similar to the recommended model processing procedure or the classified model processing procedure may be adopted to process network models such as an image segmentation model, a feature extraction model, an image recognition model, a text processing model, an audio processing model, and a video processing model so as to obtain model accuracy corresponding to the models.
In some embodiments, taking any recommendation model as an example, comparing the recommendation model processing method provided by the embodiments of the present disclosure with a method for obtaining recommendation accuracy in the related art, it can be seen that the recommendation accuracy obtained by using the method of the embodiments of the present disclosure is more accurate.
(1) Reproducible neural network search 101(NAS-Bench-101)
As can be seen from table 1, the training samples are different, and the effect of the final precision prediction model is also different, but the effect of the method (TNASP) provided by the embodiment of the present disclosure is significantly higher than that of other prior art techniques under the same number of training samples.
TABLE 1
Figure BDA0003345560400000261
Figure BDA0003345560400000271
(2) Reproducible neural network search 201(NAS-Bench-201)
As can be seen from table 2, the training samples are different, and the effect of the final precision prediction model is also different, but the effect of the method (TNASP) provided by the embodiment of the present disclosure is significantly higher than that of other prior art techniques under the same number of training samples.
TABLE 2
Training sample 780 156 469 781 1563
Neural predictor 0.343 0.413 0.584 0.634 0.646
NAO 0.467 0.493 0.470 0.522 0.526
TNASP 0.539 0.589 0.640 0.689 0.724
Neural predictor + SE 0.377 0.433 0.620 0.652 0.649
NAO+SE 0.511 0.511 0.514 0.529 0.528
TNASP+SE 0.565 0.594 0.642 0.690 0.726
(3) DARTS (Differenceble Architecture Search, micro neural network Search)
As can be seen from table 3, the method (TNASP) provided by the embodiments of the present disclosure is superior to other prior art techniques in terms of accuracy, parameters, and cost.
TABLE 3
Architecture Accuracy of measurement Parameter(s) Cost of search
DenseNet-BC 96.54 25.6 -
PyramidNet-BC 96.69 26.0 -
NASNet-A+cutout 0.539 0.589 0.640
NASNet-B 0.377 0.433 0.620
NASNet-C 0.511 0.511 0.514
AmoebaNet-A+cutout 96.66 3.2 3150
AmoebaNet-B+cutout 96.63 2.8 3150
SNAS 97.02 28.6 200
GHN 97.16 5.7 0.8
PNAS+cutout 97.17 3.2 -
DARTA+cutout 97.24 3.6 4
CTNAS+cutout 97.41 3.6 0.3
TNASP+cutout 97.48 3.7 0.3
(4)ProxylessNAS
As can be seen from table 3, the accuracy of the method (TNASP) provided by the embodiments of the present disclosure is superior to other prior art techniques.
TABLE 4
Method Parameter(s) Flip-flop First 1% The first 5%
FBNet-C 4.4 375 74.9 92.1
Proxyless(GPU) 7.0 457 75.1 92.5
SPOS 5.4 472 74.8 -
RLNAS 5.3 473 75.6 92.6
Neural predictor 6.4 536 74.75 -
NAO 6.5 590 75.5 92.5
TNASP-A 5.0 433 75.1 92.3
TNASP-B 5.1 478 75.5 92.5
TNASP-C 5.3 479 75.8 92.7
Fig. 13 is a block diagram illustrating a recommendation model processing apparatus according to an example embodiment. Referring to fig. 13, the apparatus includes:
a recommendation model obtaining unit 1301 configured to execute obtaining of a recommendation model, where the recommendation model includes multiple operators arranged in sequence, the multiple operators sequentially process account data and item data input to the recommendation model according to the sequence to obtain recommendation parameters, and the recommendation parameters indicate whether to recommend an item to an account;
a first feature obtaining unit 1302, configured to perform obtaining, based on the recommendation model, an operator feature corresponding to each operator in the recommendation model and a structural feature of the recommendation model, where the structural feature represents a connection relationship between any two operators in the multiple operators;
the second feature obtaining unit 1303 is configured to perform feature extraction based on the operator features and the structural features corresponding to each operator to obtain model features corresponding to the recommended model;
and a recommendation precision prediction unit 1304 configured to perform precision prediction based on the model features to obtain recommendation precision corresponding to the recommendation model.
According to the device provided by the embodiment of the disclosure, the operator characteristics and the structural characteristics corresponding to the recommendation model are obtained firstly, and then the operator characteristics and the structural characteristics are utilized to extract effective model characteristics, so that the characteristics of the recommendation model can be fully obtained by the mode of obtaining the model characteristics, the obtained model characteristics can more accurately represent the recommendation model, the recommendation precision obtained based on the model characteristic prediction is more accurate, and the accuracy of the recommendation precision is improved.
In some embodiments, the recommendation model further includes an index identifier corresponding to each operator, and the first feature obtaining unit 1302 includes:
the operator characteristic acquisition subunit is configured to execute query of the operator characteristic corresponding to each operator from the operator characteristic table based on the index identification corresponding to each operator;
the operator feature table comprises index identifications and operator features corresponding to at least one operator, wherein the operator features are obtained by carrying out feature extraction on the operators.
In some embodiments, the recommendation model includes N operators, where N is an integer greater than 1, and the first feature obtaining unit 1302 includes:
the structural feature obtaining subunit is configured to execute creating a target matrix with N rows and N columns based on N operators in the recommendation model, wherein the N operators are respectively used as the N rows and the N columns of the target matrix, the rows and the columns corresponding to elements on diagonal lines in the target matrix are the same operator, the value of the element on the diagonal line in the target matrix is equal to the number of other operators connected with the operator corresponding to the element, the rows and the columns corresponding to the elements on other positions in the target matrix are two different operators, and the value of the element on the other positions indicates whether the two operators corresponding to the element have a connection relation;
and the structural feature acquisition subunit is further configured to perform feature extraction on the target matrix to obtain a structural feature.
In some embodiments, the second feature obtaining unit 1303 includes:
the feature splicing subunit is configured to splice the operator features and the structural features corresponding to each operator to obtain splicing features corresponding to the recommendation model;
and the coding subunit is configured to perform coding on the splicing characteristics to obtain the model characteristics.
In some embodiments, the recommendation precision prediction model includes a feature extraction network and a prediction network,
the characteristic extraction network is used for carrying out characteristic extraction on the basis of operator characteristics and structural characteristics corresponding to each operator to obtain model characteristics;
and the prediction network is used for predicting the accuracy of the model characteristics to obtain the recommendation accuracy.
In some embodiments, the recommendation accuracy prediction model further comprises a feature acquisition network,
the characteristic acquisition network is used for acquiring operator characteristics corresponding to each operator in the recommendation model;
and the characteristic acquisition network is also used for acquiring the structural characteristics of the recommendation model.
In some embodiments, referring to fig. 14, the recommendation model processing apparatus further includes:
the model training unit 1305 is configured to execute obtaining a sample operator characteristic, a first sample structural characteristic and a first sample recommendation precision, which correspond to each sample operator in the first sample recommendation model, where the first sample structural characteristic represents a connection relationship between any two sample operators in the first sample recommendation model;
the model training unit 1305 is further configured to execute calling of a recommendation precision prediction model, and process the sample operator characteristics corresponding to each sample operator in the first sample recommendation model and the first sample structure characteristics to obtain first prediction recommendation precision;
the model training unit 1305 is further configured to perform training of the recommendation accuracy prediction model based on the first prediction recommendation accuracy and the first sample recommendation accuracy.
In some embodiments, referring to fig. 14, the recommendation model processing apparatus further includes:
a recommending unit 1306 configured to perform inputting account data and item data into a recommendation model;
the recommending unit 1306 is further configured to execute calling of a first operator in the recommending model, process the account data and the item data, and obtain output data of the first operator;
the recommending unit 1306 is further configured to perform, by taking the output data of the first operator as input data of a next operator connected to the first operator, invoking the next operator to process the output data until the output data of the last operator in the recommendation model is obtained, and determining the output data of the last operator as a recommendation parameter.
In some embodiments, the model obtaining unit 1301 is configured to perform obtaining a plurality of recommendation models, where operators in any two recommendation models are different, or connection relationships between operators in any two recommendation models are different;
referring to fig. 14, the recommendation model processing apparatus further includes:
a model selecting unit 1307 configured to execute selecting a target recommendation model corresponding to the maximum recommendation accuracy from the plurality of recommendation models based on the recommendation accuracies corresponding to the plurality of recommendation models;
and a recommending unit 1306, configured to execute calling the target recommendation model for recommendation.
In some embodiments, the recommendation model further includes a plurality of edges, each edge connecting two operators, the operators being configured to process input data input to the operators to obtain output data; and the output data of one operator connected by each edge is used as the input data of the other operator connected by each edge.
With regard to the apparatus in the above-described embodiment, the specific manner in which each unit performs the operation has been described in detail in the embodiment related to the method, and will not be described in detail here.
FIG. 15 is a block diagram illustrating an apparatus for training a precision predictive model in accordance with an exemplary embodiment. Referring to fig. 15, the apparatus includes:
the sample acquiring unit 1501 is configured to acquire a sample operator characteristic, a first sample structural characteristic and a first sample recommendation precision corresponding to each sample operator in the first sample recommendation model, wherein the first sample structural characteristic represents a connection relation between any two sample operators in the first sample recommendation model;
the precision prediction unit 1502 is configured to execute calling of a recommendation precision prediction model, perform feature extraction on the sample operator features and the first sample structure features corresponding to each sample operator in the first sample recommendation model to obtain model features corresponding to the first sample recommendation model, and perform precision prediction based on the model features corresponding to the first sample recommendation model to obtain first prediction recommendation precision;
a model training unit 1503 configured to perform training of the recommendation accuracy prediction model based on the first prediction recommendation accuracy and the first sample recommendation accuracy.
According to the device provided by the embodiment of the disclosure, the recommendation precision prediction model for predicting the recommendation precision is trained, then, the recommendation precision prediction model can be utilized to process each operator characteristic and structural characteristic corresponding to the recommendation model, and effective model characteristics are extracted by utilizing the operator characteristics and the structural characteristics, so that the characteristics of the recommendation model are fully obtained, the obtained model characteristics can more accurately represent the recommendation model, the recommendation precision obtained by the recommendation precision prediction model based on the model characteristics is more accurate, and the accuracy of the recommendation precision is improved.
In some embodiments, the sample acquiring unit 1501 is further configured to perform: obtaining sample operator characteristics, second sample structure characteristics and proxy recommendation precision corresponding to each sample operator in a second sample recommendation model, wherein the second sample structure characteristics represent a connection relation between any two sample operators in the second sample recommendation model, and the proxy recommendation precision is a variable for indicating the recommendation precision of the second sample recommendation model;
the precision prediction unit 1502 is configured to execute calling of a recommendation precision prediction model, perform feature extraction on the sample operator features and the second sample structure features corresponding to each sample operator in the second sample recommendation model to obtain model features corresponding to the second sample recommendation model, and perform precision prediction based on the model features corresponding to the second sample recommendation model to obtain second prediction recommendation precision;
a model training unit 1503 configured to perform training of the recommendation accuracy prediction model based on the first prediction recommendation accuracy, the first sample recommendation accuracy, the second prediction recommendation accuracy, and the agent recommendation accuracy.
In some embodiments, the model training unit 1503 is configured to perform:
acquiring a first error between the first prediction recommendation precision and the first sample recommendation precision and a second error between the second prediction recommendation precision and the second sample recommendation precision;
and adjusting the model parameters and the proxy recommendation precision in the recommendation precision prediction model based on the first error and the second error so that the error between the prediction recommendation precision corresponding to the historical time and the adjusted proxy recommendation precision of the second sample recommendation model approaches to the target error corresponding to the historical time.
In some embodiments, the model training unit 1503 is configured to perform:
based on the first prediction recommendation precision, the first sample recommendation precision, the second prediction recommendation precision and the agent recommendation precision, training a recommendation precision prediction model by adopting the following loss function:
Figure BDA0003345560400000311
wherein L represents a loss function, fθ(. cndot.) represents a recommended accuracy prediction model, θ represents a model parameter in the recommended accuracy prediction model, fθ(xi) Represents the first prediction recommendation precision, x, corresponding to the ith first sample recommendation modeliRepresenting sample operator characteristics and first sample structure characteristics corresponding to each sample operator in the ith first sample recommendation model, n representing the number of first sample recommendation models used for training recommendation precision prediction models, yiRepresenting a first sample recommendation precision, f, corresponding to the ith first sample recommendation modelθ(vj) Represents a second prediction recommendation precision, v, corresponding to the jth second sample recommendation modeljRepresenting the sample operator characteristic and the second sample structure characteristic corresponding to each sample operator in the jth second sample recommendation model, V representing the number of second sample recommendation models used for training the recommendation precision prediction model,
Figure BDA0003345560400000321
indicating the proxy recommendation precision corresponding to the jth second sample recommendation model,
Figure BDA0003345560400000322
representing the agent recommendation precision, and alpha is a reference parameter;
the loss function satisfies the following condition:
Figure BDA0003345560400000323
wherein V represents the number of second sample recommendation models used to train the recommendation accuracy prediction model,
Figure BDA0003345560400000324
representing the proxy recommendation precision corresponding to the jth second sample recommendation model,
Figure BDA0003345560400000325
representing the prediction recommendation precision corresponding to the jth second sample recommendation model at the tth moment, e(t)The target error corresponding to the T-th time is shown, and T shows T historical times.
With regard to the apparatus in the above-described embodiment, the specific manner in which each unit performs the operation has been described in detail in the embodiment related to the method, and will not be described in detail here.
Fig. 16 is a block diagram illustrating a classification model processing apparatus according to an exemplary embodiment. Referring to fig. 16, the apparatus includes:
the classification model obtaining unit 1601 is configured to perform obtaining of a classification model, where the classification model includes a plurality of operators arranged in sequence, and the plurality of operators sequentially process data to be classified input to the classification model according to the sequence to obtain a data type corresponding to the data;
a first feature obtaining unit 1602, configured to perform obtaining, based on the classification model, an operator feature corresponding to each operator in the classification model and a structural feature of the classification model, where the structural feature represents a connection relationship between any two operators in the plurality of operators;
a second feature obtaining unit 1603, configured to perform feature extraction based on the operator feature corresponding to each operator and the structural feature to obtain a model feature corresponding to the classification model;
a classification accuracy prediction unit 1604 configured to perform accuracy prediction based on the model features to obtain classification accuracy corresponding to the classification model.
According to the device provided by the embodiment of the disclosure, the operator characteristics and the structural characteristics corresponding to the classification model are firstly obtained, and then the operator characteristics and the structural characteristics are utilized to extract effective model characteristics, so that the characteristics of the classification model can be fully obtained by the mode of obtaining the model characteristics, the obtained model characteristics can more accurately represent the classification model, the classification precision obtained based on the model characteristic prediction is more accurate, and the accuracy of the classification precision is improved.
With regard to the apparatus in the above-described embodiment, the specific manner in which each unit performs the operation has been described in detail in the embodiment related to the method, and will not be described in detail here.
In an exemplary embodiment, an electronic device is provided that includes one or more processors, and a memory to store instructions executable by the one or more processors; wherein the one or more processors are configured to perform the recommendation model processing method, the recommendation accuracy prediction model training method, or the classification model processing method in the above embodiments.
In one possible implementation, the electronic device is provided as a terminal. Fig. 17 is a block diagram illustrating a structure of a terminal 1700 according to an example embodiment. The terminal 1700 may be a portable mobile terminal such as: a smart phone, a tablet computer, an MP3 player (Moving Picture Experts Group Audio Layer III, motion video Experts compression standard Audio Layer 3), an MP4 player (Moving Picture Experts Group Audio Layer IV, motion video Experts compression standard Audio Layer 4), a notebook computer, or a desktop computer. Terminal 1700 may also be referred to by other names such as user equipment, portable terminal, laptop terminal, desktop terminal, and the like.
The terminal 1700 includes: a processor 1701 and a memory 1702.
The processor 1701 may include one or more processing cores, such as 4-core processors, 8-core processors, and the like. The processor 1701 may be implemented in at least one hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), and a PLA (Programmable Logic Array). The processor 1701 may also include a main processor, which is a processor for Processing data in an awake state, also called a Central Processing Unit (CPU), and a coprocessor; a coprocessor is a low power processor for processing data in a standby state. In some embodiments, the processor 1701 may be integrated with a GPU (Graphics Processing Unit) that is responsible for rendering and drawing content that the display screen needs to display. In some embodiments, the processor 1701 may further include an AI (Artificial Intelligence) processor for processing computing operations related to machine learning.
The memory 1702 may include one or more computer-readable storage media, which may be non-transitory. The memory 1702 may also include high-speed random access memory, as well as non-volatile memory, such as one or more magnetic disk storage devices, flash memory storage devices. In some embodiments, a non-transitory computer readable storage medium in the memory 1702 is used to store at least one program code for execution by the processor 1701 to implement the recommendation model processing method, the recommendation accuracy prediction model training method, or the classification model processing method provided by the method embodiments of the present disclosure.
In some embodiments, terminal 1700 may also optionally include: a peripheral interface 1703 and at least one peripheral. The processor 1701, memory 1702 and peripheral interface 1703 may be connected by buses or signal lines. Various peripheral devices may be connected to peripheral interface 1703 by a bus, signal line, or circuit board. Specifically, the peripheral device includes: at least one of radio frequency circuit 1704, display screen 1705, camera assembly 1706, audio circuit 1707, positioning assembly 1708, and power supply 1709.
The peripheral interface 1703 may be used to connect at least one peripheral associated with I/O (Input/Output) to the processor 1701 and the memory 1702. In some embodiments, the processor 1701, memory 1702, and peripheral interface 1703 are integrated on the same chip or circuit board; in some other embodiments, any one or both of the processor 1701, the memory 1702, and the peripheral interface 1703 may be implemented on separate chips or circuit boards, which are not limited in this embodiment.
The Radio Frequency circuit 1704 is used for receiving and transmitting RF (Radio Frequency) signals, also called electromagnetic signals. The radio frequency circuit 1704 communicates with a communication network and other communication devices via electromagnetic signals. The rf circuit 1704 converts the electrical signal into an electromagnetic signal for transmission, or converts the received electromagnetic signal into an electrical signal. Optionally, the radio frequency circuit 1704 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a subscriber identity module card, and so forth. The radio frequency circuit 1704 may communicate with other terminals via at least one wireless communication protocol. The wireless communication protocols include, but are not limited to: the world wide web, metropolitan area networks, intranets, generations of mobile communication networks (2G, 3G, 4G, and 5G), Wireless local area networks, and/or WiFi (Wireless Fidelity) networks. In some embodiments, the radio frequency circuit 1704 may also include NFC (Near Field Communication) related circuits, which are not limited by this disclosure.
The display screen 1705 is used to display a UI (User Interface). The UI may include graphics, text, icons, video, and any combination thereof. When the display screen 1705 is a touch display screen, the display screen 1705 also has the ability to capture touch signals on or above the surface of the display screen 1705. The touch signal may be input as a control signal to the processor 1701 for processing. At this point, the display 1705 may also be used to provide virtual buttons and/or a virtual keyboard, also referred to as soft buttons and/or a soft keyboard. In some embodiments, display 1705 may be one, disposed on a front panel of terminal 1700; in other embodiments, display 1705 may be at least two, each disposed on a different surface of terminal 1700 or in a folded design; in other embodiments, display 1705 may be a flexible display disposed on a curved surface or a folded surface of terminal 1700. Even further, the display screen 1705 may be arranged in a non-rectangular irregular figure, i.e., a shaped screen. The Display screen 1705 may be made of LCD (Liquid Crystal Display), OLED (Organic Light-Emitting Diode), or the like.
The camera assembly 1706 is used to capture images or video. Optionally, camera assembly 1706 includes a front camera and a rear camera. The front camera is arranged on the front panel of the terminal, and the rear camera is arranged on the back of the terminal. In some embodiments, the number of the rear cameras is at least two, and each rear camera is any one of a main camera, a depth-of-field camera, a wide-angle camera and a telephoto camera, so that the main camera and the depth-of-field camera are fused to realize a background blurring function, and the main camera and the wide-angle camera are fused to realize panoramic shooting and VR (Virtual Reality) shooting functions or other fusion shooting functions. In some embodiments, camera assembly 1706 may also include a flash. The flash lamp can be a monochrome temperature flash lamp or a bicolor temperature flash lamp. The double-color-temperature flash lamp is a combination of a warm-light flash lamp and a cold-light flash lamp, and can be used for light compensation at different color temperatures.
The audio circuit 1707 may include a microphone and a speaker. The microphone is used for collecting sound waves of a user and the environment, converting the sound waves into electric signals, inputting the electric signals into the processor 1701 for processing, or inputting the electric signals into the radio frequency circuit 1704 for voice communication. For stereo capture or noise reduction purposes, multiple microphones may be provided, each at a different location of terminal 1700. The microphone may also be an array microphone or an omni-directional pick-up microphone. The speaker is used to convert electrical signals from the processor 1701 or the radio frequency circuit 1704 into sound waves. The loudspeaker can be a traditional film loudspeaker or a piezoelectric ceramic loudspeaker. When the speaker is a piezoelectric ceramic speaker, the speaker can be used for purposes such as converting an electric signal into a sound wave audible to a human being, or converting an electric signal into a sound wave inaudible to a human being to measure a distance. In some embodiments, the audio circuitry 1707 may also include a headphone jack.
The positioning component 1708 is used to locate the current geographic Location of the terminal 1700 to implement navigation or LBS (Location Based Service). The Positioning component 1708 may be a Positioning component based on a GPS (Global Positioning System) in the united states, a beidou System in china, a greiner Positioning System in russia, or a galileo Positioning System in the european union.
Power supply 1709 is used to power the various components in terminal 1700. The power supply 1709 may be ac, dc, disposable or rechargeable. When the power supply 1709 includes a rechargeable battery, the rechargeable battery may be a wired rechargeable battery or a wireless rechargeable battery. The wired rechargeable battery is a battery charged through a wired line, and the wireless rechargeable battery is a battery charged through a wireless coil. The rechargeable battery may also be used to support fast charge technology.
In some embodiments, terminal 1700 also includes one or more sensors 1710. The one or more sensors 1710 include, but are not limited to: acceleration sensor 1711, gyro sensor 1712, pressure sensor 1713, fingerprint sensor 1714, optical sensor 1715, and proximity sensor 1716.
The acceleration sensor 1711 can detect the magnitude of acceleration on three coordinate axes of the coordinate system established with the terminal 1700. For example, the acceleration sensor 1711 may be used to detect components of gravitational acceleration in three coordinate axes. The processor 1701 may control the display screen 1705 to display the user interface in a landscape view or a portrait view according to the gravitational acceleration signal collected by the acceleration sensor 1711. The acceleration sensor 1711 may also be used for acquisition of motion data of a game or a user.
The gyro sensor 1712 may detect a body direction and a rotation angle of the terminal 1700, and the gyro sensor 1712 may cooperate with the acceleration sensor 1711 to acquire a 3D motion of the user on the terminal 1700. The processor 1701 may perform the following functions based on the data collected by the gyro sensor 1712: motion sensing (such as changing the UI according to a user's tilting operation), image stabilization at the time of photographing, game control, and inertial navigation.
Pressure sensors 1713 may be disposed on the side frames of terminal 1700 and/or underlying display screen 1705. When the pressure sensor 1713 is disposed on the side frame of the terminal 1700, the user's grip signal to the terminal 1700 can be detected, and the processor 1701 performs left-right hand recognition or shortcut operation according to the grip signal collected by the pressure sensor 1713. When the pressure sensor 1713 is disposed below the display screen 1705, the processor 1701 controls the operability control on the UI interface according to the pressure operation of the user on the display screen 1705. The operability control comprises at least one of a button control, a scroll bar control, an icon control and a menu control.
The fingerprint sensor 1714 is configured to capture a fingerprint of the user, and the processor 1701 is configured to identify the user based on the fingerprint captured by the fingerprint sensor 1714, or the fingerprint sensor 1714 is configured to identify the user based on the captured fingerprint. Upon identifying that the user's identity is a trusted identity, the processor 1701 authorizes the user to perform relevant sensitive operations including unlocking the screen, viewing encrypted information, downloading software, paying for and changing settings, etc. Fingerprint sensor 1714 may be disposed on the front, back, or side of terminal 1700. When a physical key or vendor Logo is provided on terminal 1700, fingerprint sensor 1714 may be integrated with the physical key or vendor Logo.
The optical sensor 1715 is used to collect the ambient light intensity. In one embodiment, the processor 1701 may control the display brightness of the display screen 1705 based on the ambient light intensity collected by the optical sensor 1715. Specifically, when the ambient light intensity is high, the display brightness of the display screen 1705 is increased; when the ambient light intensity is low, the display brightness of the display screen 1705 is reduced. In another embodiment, the processor 1701 may also dynamically adjust the shooting parameters of the camera assembly 1706 according to the ambient light intensity collected by the optical sensor 1715.
A proximity sensor 1716, also known as a distance sensor, is disposed on the front panel of terminal 1700. Proximity sensor 1716 is used to gather the distance between the user and the front face of terminal 1700. In one embodiment, when proximity sensor 1716 detects that the distance between the user and the front surface of terminal 1700 is gradually reduced, processor 1701 controls display 1705 to switch from a bright screen state to a dark screen state; when proximity sensor 1716 detects that the distance between the user and the front surface of terminal 1700 is gradually increased, processor 1701 controls display 1705 to switch from the sniff state to the brighten state.
Those skilled in the art will appreciate that the architecture shown in fig. 17 is not intended to be limiting with respect to terminal 1700, and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components may be employed.
In another possible implementation, the electronic device is provided as a server. Fig. 18 is a block diagram illustrating a server 1800 that may have a large difference due to different configurations or performances according to an exemplary embodiment, and may include one or more processors (CPUs) 1801 and one or more memories 1802, where the memory 1802 stores at least one program code, and the at least one program code is loaded and executed by the processors 1801 to implement the methods provided by the above method embodiments. Of course, the server may also have components such as a wired or wireless network interface, a keyboard, and an input/output interface, so as to perform input/output, and the server may also include other components for implementing the functions of the device, which are not described herein again.
In an exemplary embodiment, a computer-readable storage medium is also provided, and when instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the steps executed by the electronic device in the recommendation model processing method, the recommendation precision prediction model training method, or the classification model processing method. Alternatively, the computer-readable storage medium may be a ROM (Read Only Memory), a RAM (Random Access Memory), a CD-ROM (Compact Disc Read-Only Memory), a magnetic tape, a floppy disk, an optical data storage device, and the like.
In an exemplary embodiment, a computer program product is also provided, which includes a computer program executed by a processor to implement the above-described recommendation model processing method, recommendation accuracy prediction model training method, or classification model processing method.
In some embodiments, the computer program according to the embodiments of the present application may be deployed to be executed on one computer device or on multiple computer devices located at one site, or may be executed on multiple computer devices distributed at multiple sites and interconnected by a communication network, and the multiple computer devices distributed at the multiple sites and interconnected by the communication network may constitute a block chain system.
Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure herein. This disclosure is intended to cover any variations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.
It will be understood that the present disclosure is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims (10)

1. A recommendation model processing method, the method comprising:
acquiring a recommendation model, wherein the recommendation model comprises a plurality of operators which are arranged in sequence, the operators sequentially process account data and article data which are input into the recommendation model according to the sequence to obtain recommendation parameters, and the recommendation parameters indicate whether the articles are recommended to the accounts or not;
acquiring operator characteristics corresponding to each operator in the recommendation model and structural characteristics of the recommendation model based on the recommendation model, wherein the structural characteristics represent the connection relation between any two operators in the multiple operators;
performing feature extraction based on the operator features corresponding to each operator and the structural features to obtain model features corresponding to the recommended models;
and performing precision prediction based on the model characteristics to obtain recommendation precision corresponding to the recommendation model.
2. The recommendation model processing method according to claim 1, wherein the recommendation model further includes an index identifier corresponding to each operator, and the step of obtaining the operator characteristic corresponding to each operator includes:
inquiring operator characteristics corresponding to each operator from an operator characteristic table based on the index identification corresponding to each operator;
the operator feature table comprises index identifications and operator features corresponding to at least one operator, wherein the operator features are obtained by performing feature extraction on the operators.
3. A recommendation accuracy prediction model training method is characterized by comprising the following steps:
obtaining a sample operator characteristic, a first sample structural characteristic and a first sample recommendation precision corresponding to each sample operator in a first sample recommendation model, wherein the first sample structural characteristic represents a connection relation between any two sample operators in the first sample recommendation model;
calling a recommendation precision prediction model, performing feature extraction on sample operator features and first sample structure features corresponding to each sample operator in the first sample recommendation model to obtain model features corresponding to the first sample recommendation model, and performing precision prediction based on the model features corresponding to the first sample recommendation model to obtain first prediction recommendation precision;
training the recommendation precision prediction model based on the first prediction recommendation precision and the first sample recommendation precision.
4. A classification model processing method, characterized in that the method comprises:
acquiring a classification model, wherein the classification model comprises a plurality of operators which are arranged in sequence, and the operators sequentially process data to be classified input into the classification model according to the sequence to obtain data types corresponding to the data;
based on the classification model, acquiring operator characteristics corresponding to each operator in the classification model and structural characteristics of the classification model, wherein the structural characteristics represent a connection relation between any two operators in the multiple operators;
performing feature extraction based on the operator features corresponding to each operator and the structural features to obtain model features corresponding to the classification models;
and performing precision prediction based on the model characteristics to obtain the classification precision corresponding to the classification model.
5. A recommendation model processing apparatus, characterized in that the apparatus comprises:
the recommendation model obtaining unit is configured to execute obtaining of a recommendation model, the recommendation model comprises a plurality of operators which are arranged in sequence, the operators sequentially process account data and article data which are input into the recommendation model according to the sequence to obtain recommendation parameters, and the recommendation parameters indicate whether the articles are recommended to the accounts or not;
a first feature obtaining unit configured to perform obtaining, based on the recommendation model, an operator feature corresponding to each operator in the recommendation model and a structural feature of the recommendation model, where the structural feature represents a connection relationship between any two operators in the plurality of operators;
the second feature acquisition unit is configured to perform feature extraction based on the operator features corresponding to each operator and the structural features to obtain model features corresponding to the recommended model;
and the recommendation precision prediction unit is configured to perform precision prediction based on the model characteristics to obtain recommendation precision corresponding to the recommendation model.
6. An apparatus for training a recommendation precision prediction model, the apparatus comprising:
the sample acquiring unit is configured to acquire a sample operator characteristic, a first sample structural characteristic and a first sample recommendation precision corresponding to each sample operator in a first sample recommendation model, wherein the first sample structural characteristic represents a connection relation between any two sample operators in the first sample recommendation model;
the precision prediction unit is configured to execute calling of a recommended precision prediction model, perform feature extraction on sample operator features and first sample structure features corresponding to each sample operator in the first sample recommended model to obtain model features corresponding to the first sample recommended model, and perform precision prediction based on the model features corresponding to the first sample recommended model to obtain first predicted recommended precision;
a model training unit configured to perform training of the recommendation accuracy prediction model based on the first prediction recommendation accuracy and the first sample recommendation accuracy.
7. A classification model processing apparatus, characterized in that the apparatus comprises:
the classification model acquisition unit is configured to execute acquisition of a classification model, the classification model comprises a plurality of operators which are arranged in sequence, and the operators sequentially process data to be classified input into the classification model according to the sequence to obtain data types corresponding to the data;
a first feature obtaining unit configured to perform obtaining, based on the classification model, an operator feature corresponding to each operator in the classification model and a structural feature of the classification model, where the structural feature represents a connection relationship between any two operators in the plurality of operators;
the second feature acquisition unit is configured to perform feature extraction based on the operator features corresponding to each operator and the structural features to obtain model features corresponding to the classification models;
and the classification precision prediction unit is configured to execute precision prediction based on the model characteristics to obtain the classification precision corresponding to the classification model.
8. An electronic device, characterized in that the electronic device comprises:
one or more processors;
a memory for storing the one or more processor-executable instructions;
wherein the one or more processors are configured to perform the recommendation model processing method of any one of claims 1-2, or configured to perform the recommendation accuracy prediction model training method of claim 3, or configured to perform the classification model processing method of claim 4.
9. A computer-readable storage medium, wherein instructions in the computer-readable storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the recommendation model processing method of any one of claims 1 to 2, or to perform the recommendation accuracy prediction model training method of claim 3, or to perform the classification model processing method of claim 4.
10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the recommendation model processing method of any one of claims 1 to 2, or implements the recommendation accuracy prediction model training method of claim 3, or implements the classification model processing method of claim 4.
CN202111320914.2A 2021-11-09 2021-11-09 Recommendation model processing method and device, electronic equipment and storage medium Active CN114117206B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111320914.2A CN114117206B (en) 2021-11-09 2021-11-09 Recommendation model processing method and device, electronic equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111320914.2A CN114117206B (en) 2021-11-09 2021-11-09 Recommendation model processing method and device, electronic equipment and storage medium

Publications (2)

Publication Number Publication Date
CN114117206A true CN114117206A (en) 2022-03-01
CN114117206B CN114117206B (en) 2023-05-30

Family

ID=80378019

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111320914.2A Active CN114117206B (en) 2021-11-09 2021-11-09 Recommendation model processing method and device, electronic equipment and storage medium

Country Status (1)

Country Link
CN (1) CN114117206B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114547428A (en) * 2022-03-07 2022-05-27 北京达佳互联信息技术有限公司 Recommendation model processing method and device, electronic equipment and storage medium
CN115167846A (en) * 2022-09-07 2022-10-11 浙江大华技术股份有限公司 Recommendation method of downstream operator, electronic device and computer-readable storage medium

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110378413A (en) * 2019-07-17 2019-10-25 Oppo广东移动通信有限公司 Neural network model processing method, device and electronic equipment
CN111612134A (en) * 2020-05-20 2020-09-01 鼎富智能科技有限公司 Neural network structure searching method and device, electronic equipment and storage medium
CN111738098A (en) * 2020-05-29 2020-10-02 浪潮(北京)电子信息产业有限公司 Vehicle identification method, device, equipment and storage medium
CN112508177A (en) * 2020-12-11 2021-03-16 鼎富智能科技有限公司 Network structure searching method and device, electronic equipment and storage medium
CN112784949A (en) * 2021-01-28 2021-05-11 华东计算技术研究所（中国电子科技集团公司第三十二研究所） Neural network architecture searching method and system based on evolutionary computation
CN113361680A (en) * 2020-03-05 2021-09-07 华为技术有限公司 Neural network architecture searching method, device, equipment and medium
US20210312261A1 (en) * 2020-04-02 2021-10-07 Huawei Technologies Co., Ltd. Neural network search method and related apparatus
CN113590968A (en) * 2021-08-10 2021-11-02 平安普惠企业管理有限公司 Resource recommendation method and device, computer equipment and storage medium

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110378413A (en) * 2019-07-17 2019-10-25 Oppo广东移动通信有限公司 Neural network model processing method, device and electronic equipment
CN113361680A (en) * 2020-03-05 2021-09-07 华为技术有限公司 Neural network architecture searching method, device, equipment and medium
US20210312261A1 (en) * 2020-04-02 2021-10-07 Huawei Technologies Co., Ltd. Neural network search method and related apparatus
CN111612134A (en) * 2020-05-20 2020-09-01 鼎富智能科技有限公司 Neural network structure searching method and device, electronic equipment and storage medium
CN111738098A (en) * 2020-05-29 2020-10-02 浪潮(北京)电子信息产业有限公司 Vehicle identification method, device, equipment and storage medium
CN112508177A (en) * 2020-12-11 2021-03-16 鼎富智能科技有限公司 Network structure searching method and device, electronic equipment and storage medium
CN112784949A (en) * 2021-01-28 2021-05-11 华东计算技术研究所（中国电子科技集团公司第三十二研究所） Neural network architecture searching method and system based on evolutionary computation
CN113590968A (en) * 2021-08-10 2021-11-02 平安普惠企业管理有限公司 Resource recommendation method and device, computer equipment and storage medium

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
周鹏 等: "采用神经网络架构搜索的遥感影像分割方法", 《西安电子科技大学学报》 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114547428A (en) * 2022-03-07 2022-05-27 北京达佳互联信息技术有限公司 Recommendation model processing method and device, electronic equipment and storage medium
CN115167846A (en) * 2022-09-07 2022-10-11 浙江大华技术股份有限公司 Recommendation method of downstream operator, electronic device and computer-readable storage medium
CN115167846B (en) * 2022-09-07 2022-12-20 浙江大华技术股份有限公司 Recommendation method of downstream operator, electronic device and computer-readable storage medium
WO2024051146A1 (en) * 2022-09-07 2024-03-14 Zhejiang Dahua Technology Co., Ltd. Methods, systems, and computer-readable media for recommending downstream operator

Also Published As

Publication number Publication date
CN114117206B (en) 2023-05-30

Similar Documents

Publication Publication Date Title
CN108629747B (en) Image enhancement method and device, electronic equipment and storage medium
WO2020253657A1 (en) Video clip positioning method and apparatus, computer device, and storage medium
CN111182453B (en) Positioning method, positioning device, electronic equipment and storage medium
CN110134804B (en) Image retrieval method, device and storage medium
CN110471858B (en) Application program testing method, device and storage medium
CN111737573A (en) Resource recommendation method, device, equipment and storage medium
CN111127509B (en) Target tracking method, apparatus and computer readable storage medium
CN114547428A (en) Recommendation model processing method and device, electronic equipment and storage medium
CN111680697B (en) Method, device, electronic equipment and medium for realizing field adaptation
CN111897996A (en) Topic label recommendation method, device, equipment and storage medium
CN111611490A (en) Resource searching method, device, equipment and storage medium
CN114117206B (en) Recommendation model processing method and device, electronic equipment and storage medium
CN111738365B (en) Image classification model training method and device, computer equipment and storage medium
CN113918767A (en) Video clip positioning method, device, equipment and storage medium
CN109961802B (en) Sound quality comparison method, device, electronic equipment and storage medium
CN110166275B (en) Information processing method, device and storage medium
CN111611414B (en) Vehicle searching method, device and storage medium
CN112053360A (en) Image segmentation method and device, computer equipment and storage medium
CN111782950A (en) Sample data set acquisition method, device, equipment and storage medium
CN111563201A (en) Content pushing method, device, server and storage medium
CN113343709B (en) Method for training intention recognition model, method, device and equipment for intention recognition
CN113032560B (en) Sentence classification model training method, sentence processing method and equipment
CN112560612B (en) System, method, computer device and storage medium for determining business algorithm
CN114817709A (en) Sorting method, device, equipment and computer readable storage medium
CN113139614A (en) Feature extraction method and device, electronic equipment and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant