CN114663143B - Intervention user screening method and device based on differential intervention response model - Google Patents
Intervention user screening method and device based on differential intervention response model Download PDFInfo
- Publication number
- CN114663143B CN114663143B CN202210279004.2A CN202210279004A CN114663143B CN 114663143 B CN114663143 B CN 114663143B CN 202210279004 A CN202210279004 A CN 202210279004A CN 114663143 B CN114663143 B CN 114663143B
- Authority
- CN
- China
- Prior art keywords
- intervention
- characteristic information
- user
- model
- users
- 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.)
- Active
Links
- 230000004044 response Effects 0.000 title claims abstract description 80
- 238000000034 method Methods 0.000 title claims abstract description 64
- 238000012216 screening Methods 0.000 title claims abstract description 43
- 238000012549 training Methods 0.000 claims abstract description 132
- 238000003062 neural network model Methods 0.000 claims abstract description 30
- 238000012952 Resampling Methods 0.000 claims abstract description 23
- 238000012545 processing Methods 0.000 claims abstract description 14
- 238000012163 sequencing technique Methods 0.000 claims abstract description 12
- 238000006243 chemical reaction Methods 0.000 claims description 40
- 238000004891 communication Methods 0.000 claims description 32
- 238000007477 logistic regression Methods 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 12
- 238000013507 mapping Methods 0.000 claims description 11
- 238000010219 correlation analysis Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 6
- 238000013473 artificial intelligence Methods 0.000 abstract description 8
- 230000009466 transformation Effects 0.000 abstract description 8
- 238000005516 engineering process Methods 0.000 description 9
- 230000008901 benefit Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000007726 management method Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000014759 maintenance of location Effects 0.000 description 3
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 208000003443 Unconsciousness Diseases 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000011022 operating instruction Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0203—Market surveys; Market polls
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- Physics & Mathematics (AREA)
- Development Economics (AREA)
- Accounting & Taxation (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses an intervention user screening method and device based on a differential intervention response model, relates to the field of artificial intelligence, and mainly aims to solve the problem that the intervention cost is increased because a natural transformation user is judged as a target intervention user and is intervened when a target intervention client is screened only according to the transformation probability of the user. Comprising the following steps: acquiring characteristic information of a plurality of groups of users to be intervened; performing prediction processing on multiple groups of characteristic information based on a differential intervention response model which is trained by the model, obtaining multiple groups of intervention gain scores, wherein the differential intervention response model is constructed based on a multi-task learning neural network model, and model training is completed by utilizing training samples obtained by matching a resampling algorithm based on a tendency score; and sequencing the intervention gain scores of a plurality of groups, selecting users to be intervened corresponding to the intervention gain scores of a preset number as target intervention users, and executing intervention operation on the target intervention users.
Description
Technical Field
The application relates to the technical field of artificial intelligence, in particular to an intervention user screening method and device based on a differential intervention response model.
Background
With the development of big data and artificial intelligence, accurate marketing models have been applied to various stages of customer lifecycle management. At different stages of the customer life cycle, merchants typically use some intervention means to intervene in the customer, causing the customer to complete conversions of payment, upgrades, renewal, etc.
In the prior art, when a target intervention client is confirmed, a marketing model is established to predict the conversion probability of a dry and wet client, and the client with higher conversion probability is taken as the target intervention client to intervene.
However, in practical applications, clients can be classified into four types, i.e., natural conversion clients, sensitive clients, unconscious clients and reactive clients, according to the conversion conditions of clients before and after intervention. The natural transformation clients are clients which can be transformed whether to intervene or not, so the meaning of whether to intervene or not is not great for the natural transformation clients. Since the prior art screens target intervention clients only according to the conversion probability of clients, it is caused that natural conversion clients are also determined as target intervention clients and intervene, resulting in an increase in intervention cost.
Disclosure of Invention
In view of the above, the present application provides an intervention user screening method and device based on a differential intervention response model, and aims to improve the problem of increased intervention cost caused by determining a natural transformation user as a target intervention user and performing intervention when screening a target intervention client only according to a transformation probability of the user.
According to one aspect of the present application, there is provided an intervention user screening method based on a differential intervention response model, including:
Acquiring characteristic information of a plurality of groups of users to be intervened;
performing prediction processing on multiple groups of characteristic information based on a differential intervention response model which is trained by the model, obtaining multiple groups of intervention gain scores, wherein the differential intervention response model is constructed based on a multi-task learning neural network model, and model training is completed by utilizing training samples obtained by matching a resampling algorithm based on a tendency score;
and sequencing the intervention gain scores of a plurality of groups, selecting users to be intervened corresponding to the intervention gain scores of a preset number as target intervention users, and executing intervention operation on the target intervention users.
Preferably, before the predicting process is performed on the multiple sets of the feature information by the differential intervention response model based on the completed model training, the method further includes:
Constructing a multitask learning neural network model;
Initializing the multi-task learning neural network model to obtain an initial differential intervention response model;
Calculating conversion probability prediction loss parameters of a first training sample based on a loss function, and updating the initial differential intervention response model by combining a back propagation algorithm and a random gradient descent algorithm, wherein the first training sample is obtained based on a tendency score matching resampling algorithm;
And if the conversion probability prediction loss parameters meet the preset standard, completing model training to obtain a differential intervention response model for completing model training.
Preferably, before the calculating the conversion probability prediction loss parameter of the first training sample based on the loss function, the method further includes:
Respectively acquiring characteristic information of the intervention group users and characteristic information of the comparison group users;
Screening intervention characteristic information from the characteristic information of the intervention group user and the characteristic information of the control group user based on a correlation analysis algorithm to obtain a second training sample, wherein the second training sample is used for representing training samples for training a two-classification logistic regression model;
predicting a third training sample based on the two-class logistic regression model with the model training completed to obtain a plurality of intervention group user tendency scores and a plurality of control group user tendency scores, wherein the third training sample is used for representing the characteristic information of the users except the second training sample.
Preferably, after predicting the third training sample based on the two-classification logistic regression model trained by the completed model to obtain the user tendency scores of the plurality of intervention groups and the user tendency scores of the plurality of control groups, the method further comprises:
Sorting the user trend scores of the intervention group to obtain a user trend score sequence of the intervention group, and grouping the user trend score sequence of the intervention group by combining an equal frequency bin algorithm to obtain a plurality of trend score intervals of the intervention group;
screening the characteristic information of the control group user matched with each intervention group trend score interval from the characteristic information of the control group user;
And combining the characteristic information of the intervention group users and the characteristic information of the control group users in the same trend score interval according to a preset proportion to generate a first training sample.
Preferably, the performing an intervention operation on the target intervention user specifically includes:
Pre-constructing a mapping relation between the intervention operation category and the characteristic information of the user;
determining an intervention operation category matched with the characteristic information of the target intervention user based on the mapping relation;
And calling an intervention operation thread matched with the intervention operation category, and executing intervention operation on the target intervention user based on the intervention operation thread.
Preferably, after the performing, based on the intervention operation thread, an intervention operation on the target intervention user, the method further includes:
and periodically acquiring intervention operation feedback information according to a preset time interval to serve as a basis for updating the intervention operation category.
Preferably, before the obtaining the characteristic information of the multiple groups of users to be intervened, the method further includes:
Acquiring communication addresses of all users in a database;
sending a registration invitation message of a target item to the communication address;
and receiving registration return information of the full-quantity users, and storing the registration return information in a user database of the target item to serve as characteristic information of the users to be intervened.
According to another aspect of the present application, there is provided an intervention user screening apparatus based on a differential intervention response model, comprising:
The first acquisition module is used for acquiring characteristic information of a plurality of groups of users to be interfered;
The first prediction module is used for performing prediction processing on a plurality of groups of characteristic information based on a differential intervention response model which is trained by the model, so as to obtain a plurality of groups of intervention gain scores, wherein the differential intervention response model is constructed based on a multi-task learning neural network model, and training samples obtained by matching a resampling algorithm based on a tendency score are used for completing model training;
the selecting module is used for sorting the intervention gain scores of a plurality of groups, selecting users to be intervened corresponding to the intervention gain scores of a preset number as target intervention users, and executing intervention operation on the target intervention users.
Preferably, before the first prediction module, the apparatus further includes:
The construction module is used for constructing a multi-task learning neural network model;
The initialization module is used for initializing the multi-task learning neural network model to obtain an initial differential intervention response model;
The updating module is used for calculating conversion probability prediction loss parameters of a first training sample based on a loss function, updating the initial differential intervention response model by combining a back propagation algorithm and a random gradient descent algorithm, wherein the first training sample is obtained by matching a resampling algorithm based on a tendency score;
and the updating module is also used for completing model training if the conversion probability prediction loss parameter meets a preset standard, and obtaining a differential intervention response model for completing model training.
Preferably, before the updating module, the apparatus further includes:
the second acquisition module is used for respectively acquiring the characteristic information of the intervention group user and the characteristic information of the comparison group user;
The first screening module is used for screening the intervention characteristic information from the characteristic information of the intervention group user and the characteristic information of the control group user based on a correlation analysis algorithm to obtain a second training sample, and the second training sample is used for representing a training sample for training a classification logistic regression model;
the second prediction module is used for predicting a third training sample based on a two-class logistic regression model which is trained by the completed model to obtain a plurality of intervention group user tendency scores and a plurality of comparison group user tendency scores, and the third training sample is used for representing the characteristic information of the users except the second training sample.
Preferably, after the second prediction module, the apparatus further includes:
The grouping module is used for sequencing the plurality of intervention group user trend scores to obtain an intervention group user trend score sequence, and grouping the intervention group user trend score sequence by combining an equal frequency bin algorithm to obtain a plurality of intervention group trend score intervals;
The second screening module is used for screening the characteristic information of the comparison group users matched with each intervention group trend score interval from the characteristic information of the comparison group users;
and the merging module is used for merging the characteristic information of the intervention group users and the characteristic information of the comparison group users in the same trend score interval according to a preset proportion to generate a first training sample.
Preferably, the selecting module specifically includes:
the construction unit is used for pre-constructing the mapping relation between the intervention operation category and the characteristic information of the user;
the determining unit is used for determining an intervention operation category matched with the characteristic information of the target intervention user based on the mapping relation;
And the execution unit is used for calling the intervention operation thread matched with the intervention operation category and executing the intervention operation on the target intervention user based on the intervention operation thread.
Preferably, after the execution unit, the module further includes:
The acquisition unit is used for periodically acquiring the intervention operation feedback information according to the preset time interval so as to be used as a basis for updating the intervention operation category.
Preferably, before the first obtaining module, the apparatus further includes:
The third acquisition module is used for acquiring the communication addresses of the total users in the database;
A sending module, configured to send a registration invitation message of a target item to the communication address;
And the storage module is used for receiving the registration return information of the total users and storing the registration return information in a user database of the target item to be used as the characteristic information of the users to be intervened.
According to yet another aspect of the present application, there is provided a storage medium having stored therein at least one executable instruction for causing a processor to perform operations corresponding to the intervention user screening method based on a differential intervention response model as described above.
According to still another aspect of the present application, there is provided a computer apparatus including: the device comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete communication with each other through the communication bus;
the memory is configured to store at least one executable instruction, where the executable instruction causes the processor to execute operations corresponding to the intervention user screening method based on the differential intervention response model.
By means of the technical scheme, the technical scheme provided by the embodiment of the application has at least the following advantages:
The application provides an intervention user screening method and device based on a differential intervention response model, which comprises the steps of firstly obtaining characteristic information of a plurality of groups of users to be intervened; secondly, carrying out prediction processing on a plurality of groups of characteristic information based on a differential intervention response model which is trained by the model, obtaining a plurality of groups of intervention gain scores, wherein the differential intervention response model is constructed based on a multi-task learning neural network model, and training the model by utilizing a training sample obtained by matching a resampling algorithm based on a tendency score; and finally, sequencing a plurality of groups of intervention gain scores, selecting users to be intervened corresponding to the intervention gain scores in a preset number as target intervention users, and executing intervention operation on the target intervention users. Compared with the prior art, the method and the device have the advantages that the model construction is based on the multitask learning neural network, the differential intervention response model is obtained by training the training sample obtained based on the tendency score matching resampling algorithm, and the model prediction precision and the confidence level are improved; the method further utilizes the feature information of the user to be interfered to predict operation to obtain an interference gain score, screens target interference users according to the interference gain score, accurately screens target interference users sensitive to interference operation reaction, avoids the situation that a natural conversion client is judged to be a target interference client, and reduces interference cost.
The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present application more readily apparent.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the application. Also, like reference numerals are used to designate like parts throughout the figures. In the drawings:
FIG. 1 shows a flowchart of an intervention user screening method based on a differential intervention response model, provided by an embodiment of the application;
FIG. 2 illustrates a schematic diagram of the relationship between customer classification and intervention provided by an embodiment of the present application;
FIG. 3 is a schematic diagram of a multi-task neural network model according to an embodiment of the present application;
FIG. 4 shows a block diagram of an intervention user screening device based on a differential intervention response model according to an embodiment of the present application;
fig. 5 shows a schematic structural diagram of a computer device according to an embodiment of the present application.
Detailed Description
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
The embodiment of the application can acquire and process the related data based on the artificial intelligence technology. Wherein artificial intelligence (ARTIFICIAL INTELLIGENCE, AI) is the theory, method, technique, and application system that uses a digital computer or a digital computer-controlled machine to simulate, extend, and expand human intelligence, sense the environment, acquire knowledge, and use knowledge to obtain optimal results.
Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and other directions.
In this regard, in one embodiment, as shown in fig. 1, an intervention user screening method based on a differential intervention response model is provided, and the method is described by taking application to computer devices such as a server, where the server may be an independent server, or may be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligent platforms, such as an intelligent medical system, a digital medical platform, and the like. The method comprises the following steps:
101. And acquiring characteristic information of a plurality of groups of users to be intervened.
In the embodiment of the application, the current execution body can be applied to a client management system, such as a client management system of an insurance company. It will be appreciated that merchants typically intervene with users in order to cause them to perform conversions (e.g., pay, upgrade, renewal, etc.) that are expected to be completed by the users, such as, for example, insurance companies causing users to bind cards to achieve automatic renewal after expiration of insurance, or take-away platforms causing users to purchase members and sending members to users to exempt from red packs. Wherein, the characteristic information of the user is used for characterizing the characteristic information related to the intervention operation, including but not limited to basic information of the user, user portraits, historical behavior records and the like.
102. And predicting the characteristic information of the plurality of groups based on the differential intervention response model which is trained by the model, and obtaining the intervention gain scores of the plurality of groups.
The differential intervention response model is constructed based on a multi-task learning neural network model, and model training is completed by using training samples obtained based on a tendency score matching resampling algorithm. It will be appreciated that the intervention gain score is used to characterize the difference in probability of user conversion before and after intervention. In the embodiment of the application, the model is constructed based on the multi-task learning neural network, the feature representation of the bottom layer of the neural network model can be shared, the combined training of two branches of the differential intervention response model is realized, and compared with the training method for separately training the prediction model by respectively utilizing two training samples, the prediction precision of the differential intervention response model is effectively improved. In addition, the model training is carried out by utilizing the training sample obtained based on the tendency score matching resampling method, so that the distribution deviation of the sample can be effectively eliminated, and the prediction precision of the intervention response model is further improved.
103. And sequencing the intervention gain scores of the multiple groups, selecting users to be intervened corresponding to the intervention gain scores of the preset number as target intervention users, and executing intervention operation on the target intervention users.
In the embodiment of the present application, the intervention gain scores obtained in step 102 are ordered according to a preset rule, and may be arranged in descending order or ascending order. Selecting a preset number of users to be interfered as target interference users from the sorting to execute the interference operation, wherein the preset number can be that a certain number of users are selected as target interference users, for example, in the descending order, the first 50 users are selected as target interference users; or selecting the user with the intervention gain score exceeding the preset threshold value of the intervention gain score as the target intervention user, and the application is not particularly limited.
In practical application, as shown in fig. 2, clients can be classified into four types, namely, natural conversion clients, sensitive clients, non-active clients and reactive clients according to the conversion conditions of clients before and after intervention. The "natural transformation user" is a client that transforms whether or not to intervene, and therefore, the meaning of whether or not to intervene is not great for the "natural transformation user". While for "sensitive users", the impact of the intervention operation on the "sensitive users" is relatively large, so in order to save the intervention cost, the intervention operation should be performed on the "sensitive users" first, so as to improve the success efficiency and return on investment of the intervention strategy.
Compared with the prior art, the method and the device are constructed based on the multitask learning neural network model, and the differential intervention response model is obtained by training the training sample obtained based on the tendency score matching resampling algorithm, so that the progress and the confidence of model prediction are improved; the method further utilizes the feature information of the user to be interfered to predict operation to obtain an interference gain score, screens target interference users according to the interference gain score, accurately screens target interference users sensitive to interference operation reaction, avoids the situation that a natural conversion client is judged to be a target interference client, and reduces interference cost.
For further explanation and limitation, in the embodiment of the present application, before performing prediction processing on multiple sets of feature information based on the differential intervention response model that has completed model training, the method of the embodiment further includes: constructing a multitask learning neural network model; initializing the multi-task learning neural network model to obtain an initial differential intervention response model; calculating conversion probability prediction loss parameters of a first training sample based on a loss function, and updating an initial differential intervention response model by combining a random gradient descent algorithm through a back propagation algorithm; and if the conversion probability prediction loss parameters meet the preset standards, completing model training to obtain a differential intervention response model for completing model training.
The first training sample is obtained based on a tendency score matching resampling algorithm. First, a multi-tasking neural network model is constructed as shown in fig. 3. The sharing module may be a three-layer full-connection layer network, and the expert network module 0 and the expert network module 1 are two-layer full-connection layer neural networks respectively. Secondly, in order to improve the prediction precision of the differential intervention response model, the built multi-task neural network model is initialized, for example, an Xavier method can be adopted to initialize the model, and an initial differential intervention response model is obtained. And calculating a conversion probability prediction loss parameter of the first training sample based on the loss function, namely, predicting deviation between the conversion probability and the real conversion probability. In order to eliminate the deviation between the intervention group and the control group in the training samples, preferably, a tendency score matching resampling method may be used to obtain a first training sample, where the loss function is as follows:
Wherein n represents the number of samples, alpha represents the weight of l 2 norm, beta represents the ratio of the control group to the intervention group samples in the training samples, T is 1, the samples belong to the intervention group, 0 belongs to the control group, y 0,y1 respectively represents the present true values of the samples x 0,x1 respectively belong to the control group, The final output values of the expert network module 0 and the expert network module 1 as in fig. 3 represent the predicted transition probabilities of the differential intervention response models before and after the intervention, respectively. And updating the initial differential intervention response model by back propagation and combining a random gradient descent method, and when the conversion probability prediction loss parameter is smaller than a preset conversion probability prediction loss parameter threshold or the number of iterations is reached, indicating that model training is completed, and obtaining a final differential intervention response model.
For further explanation and limitation, in an embodiment of the present application, before calculating the conversion probability prediction loss parameter of the first training sample based on the loss function, the method of the embodiment further includes: respectively acquiring characteristic information of the intervention group users and characteristic information of the comparison group users; screening intervention characteristic information from the characteristic information of the intervention group user and the characteristic information of the comparison group user based on a correlation analysis algorithm to obtain a second training sample; and predicting a third training sample based on the two-class logistic regression model with the completed model training to obtain a plurality of intervention group user tendency scores and a plurality of control group user tendency scores.
The second training sample is used for representing a training sample for training the two-classification logistic regression model; the third training sample is used to characterize the characteristic information of the user other than the second training sample. Illustratively, the full-scale information of the pre-group users and the full-scale information (X, T) - > Y of the control group users are dried from the database, wherein X represents characteristic information of clients, T represents whether intervention is performed, and Y represents whether conversion is finally performed. Taking an intervention measure for prompting the user to expire the automatic renewal of the binding card in the insurance user retention management as an example, the intervention group user is a bound card user, and the contrast group user is an unbound card user. In order to sample and screen out the characteristic information of the user relevant to T from the characteristic information X of the user to obtain a second training sample for training the classification logistic regression model, the characteristic information of the user can be subjected to correlation analysis with the intervention strategy, and the characteristic information of the user relevant to T can be screened out by combining with the service requirement. Further, a two-class logistic regression model F (X) - > P is built with T as a target, and trend scores of each sample are predicted on X through F (X), and it is to be noted that the samples include feature information of the intervention group user and feature information of the control group user, and are a residual sample set after the second training sample is removed from X.
For further explanation and limitation, in the embodiment of the present application, the method further includes, after predicting the third training sample based on the two-classification logistic regression model that has completed the model training to obtain the user tendency scores of the plurality of intervention groups and the user tendency scores of the plurality of control groups: sequencing the user trend scores of the intervention group to obtain an intervention group user trend score sequence, and grouping the intervention group user trend score sequence by combining an equal frequency bin algorithm to obtain a plurality of intervention group trend score intervals; screening the characteristic information of the control group user matched with the trend score intervals of each intervention group from the characteristic information of the control group user; and combining the characteristic information of the intervention group users and the characteristic information of the control group users in the same trend score interval according to a preset proportion to generate a first training sample.
It should be noted that, in the embodiment of the present application, by determining the trend partition of each set of samples of the intervention group, further obtaining the corresponding samples of the control group from the samples of the control group, and synthesizing the training samples (i.e., the first training samples) of the differential intervention response model according to the preset proportion, the distribution deviation of the training samples is eliminated, and the prediction accuracy of the differential intervention response model is effectively improved. In addition, when the corresponding control group samples are obtained, the control group samples can be obtained in a downsampling mode under the condition that the quantity of the control group samples is sufficient; for the case where the number of control samples is less than the desired number, the control samples may be obtained in an oversampling manner.
For further explanation and limitation, in the embodiment of the present application, performing an intervention operation on a target intervention user specifically includes: pre-constructing a mapping relation between the intervention operation category and the characteristic information of the user; determining intervention operation categories matched with characteristic information of target intervention users based on the mapping relation; and calling the intervention operation thread matched with the intervention operation category, and executing the intervention operation on the target intervention user based on the intervention operation thread.
Taking insurance user retention management as an example, feature information of the user A is analyzed to obtain that the user A needs to be subjected to retention promotion, the intervention operation type of the user A needs to be subjected to card binding promotion based on a pre-constructed mapping relation, an operation thread for promoting the user to bind the card is further called, and card binding promotion intervention operation is further completed.
Optionally, in an embodiment of the present application, after performing, based on the intervention operation thread, an intervention operation on the target intervention user, the method further includes: and periodically acquiring intervention operation feedback information according to a preset time interval to serve as a basis for updating the intervention operation category.
It should be noted that, after the intervention operation is performed, feedback information after the intervention operation may be obtained to determine whether the current effect of the intervention operation is obvious, so as to serve as a basis for updating the type of the intervention operation.
Preferably, in the embodiment of the present application, before acquiring the feature information of the multiple groups of users to be intervened, the method further includes: acquiring communication addresses of all users in a database; sending a registration invitation message of the target item to the communication address; and receiving registration return information of the full-quantity users, and storing the registration return information in a user database of the target project to serve as characteristic information of the users to be intervened.
Specifically, when a new project is released, the user can be invited to join in the new project by sending registration invitation information to the existing full-quantity user, relevant information input during user registration is received and stored in a user database of the new project, and the relevant information is used as characteristic information of the user to be intervened, so that the relevant information is directly called when target intervention user screening is needed.
The application provides an intervention user screening method based on a differential intervention response model, which comprises the steps of firstly obtaining characteristic information of a plurality of groups of users to be intervened; secondly, carrying out prediction processing on a plurality of groups of characteristic information based on a differential intervention response model which is trained by the model, obtaining a plurality of groups of intervention gain scores, wherein the differential intervention response model is constructed based on a multi-task learning neural network model, and training the model by utilizing a training sample obtained by matching a resampling algorithm based on a tendency score; and finally, sequencing a plurality of groups of intervention gain scores, selecting users to be intervened corresponding to the intervention gain scores in a preset number as target intervention users, and executing intervention operation on the target intervention users. Compared with the prior art, the method and the device are constructed based on the multitask learning neural network model, and the differential intervention response model is obtained by training the training sample obtained based on the tendency score matching resampling algorithm, so that the progress and the confidence of model prediction are improved; the method further utilizes the feature information of the user to be interfered to predict operation to obtain an interference gain score, screens target interference users according to the interference gain score, accurately screens target interference users sensitive to interference operation reaction, avoids the situation that a natural conversion client is judged to be a target interference client, and reduces interference cost.
Further, as an implementation of the method shown in fig. 1, an embodiment of the present application provides an intervention user screening apparatus based on a differential intervention response model, as shown in fig. 4, where the apparatus includes:
the first obtaining module 21, the first predicting module 22 and the selecting module 23.
A first obtaining module 21, configured to obtain characteristic information of a plurality of groups of users to be intervened;
The first prediction module 22 is configured to perform prediction processing on multiple sets of the feature information based on a differential intervention response model that has been trained by the model, so as to obtain multiple sets of intervention gain scores, where the differential intervention response model is constructed based on a multi-task learning neural network model, and complete model training by using training samples obtained by matching a resampling algorithm based on a tendency score;
The selecting module 23 is configured to sort the multiple groups of intervention gain scores, select a user to be intervened corresponding to a preset number of the intervention gain scores as a target intervention user, and execute an intervention operation on the target intervention user.
In a specific application scenario, before the first prediction module, the apparatus further includes:
The construction module is used for constructing a multi-task learning neural network model;
The initialization module is used for initializing the multi-task learning neural network model to obtain an initial differential intervention response model;
The updating module is used for calculating conversion probability prediction loss parameters of a first training sample based on a loss function, updating the initial differential intervention response model by combining a back propagation algorithm and a random gradient descent algorithm, wherein the first training sample is obtained by matching a resampling algorithm based on a tendency score;
and the updating module is also used for completing model training if the conversion probability prediction loss parameter meets a preset standard, and obtaining a differential intervention response model for completing model training.
In a specific application scenario, before the updating module, the apparatus further includes:
the second acquisition module is used for respectively acquiring the characteristic information of the intervention group user and the characteristic information of the comparison group user;
The first screening module is used for screening the intervention characteristic information from the characteristic information of the intervention group user and the characteristic information of the control group user based on a correlation analysis algorithm to obtain a second training sample, and the second training sample is used for representing a training sample for training a classification logistic regression model;
the second prediction module is used for predicting a third training sample based on a two-class logistic regression model which is trained by the completed model to obtain a plurality of intervention group user tendency scores and a plurality of comparison group user tendency scores, and the third training sample is used for representing the characteristic information of the users except the second training sample.
In a specific application scenario, after the second prediction module, the apparatus further includes:
The grouping module is used for sequencing the plurality of intervention group user trend scores to obtain an intervention group user trend score sequence, and grouping the intervention group user trend score sequence by combining an equal frequency bin algorithm to obtain a plurality of intervention group trend score intervals;
The second screening module is used for screening the characteristic information of the comparison group users matched with each intervention group trend score interval from the characteristic information of the comparison group users;
and the merging module is used for merging the characteristic information of the intervention group users and the characteristic information of the comparison group users in the same trend score interval according to a preset proportion to generate a first training sample.
In a specific application scenario, the selecting module specifically includes:
the construction unit is used for pre-constructing the mapping relation between the intervention operation category and the characteristic information of the user;
the determining unit is used for determining an intervention operation category matched with the characteristic information of the target intervention user based on the mapping relation;
And the execution unit is used for calling the intervention operation thread matched with the intervention operation category and executing the intervention operation on the target intervention user based on the intervention operation thread.
In a specific application scenario, after the execution unit, the module further includes:
The acquisition unit is used for periodically acquiring the intervention operation feedback information according to the preset time interval so as to be used as a basis for updating the intervention operation category.
In a specific application scenario, before the first obtaining module, the apparatus further includes:
The third acquisition module is used for acquiring the communication addresses of the total users in the database;
A sending module, configured to send a registration invitation message of a target item to the communication address;
And the storage module is used for receiving the registration return information of the total users and storing the registration return information in a user database of the target item to be used as the characteristic information of the users to be intervened.
The application provides an intervention user screening device based on a differential intervention response model, which comprises the following steps of firstly acquiring characteristic information of a plurality of groups of users to be intervened; secondly, carrying out prediction processing on a plurality of groups of characteristic information based on a differential intervention response model which is trained by the model, obtaining a plurality of groups of intervention gain scores, wherein the differential intervention response model is constructed based on a multi-task learning neural network model, and training the model by utilizing a training sample obtained by matching a resampling algorithm based on a tendency score; and finally, sequencing a plurality of groups of intervention gain scores, selecting users to be intervened corresponding to the intervention gain scores in a preset number as target intervention users, and executing intervention operation on the target intervention users. Compared with the prior art, the method and the device are constructed based on the multitask learning neural network model, and the differential intervention response model is obtained by training the training sample obtained based on the tendency score matching resampling algorithm, so that the progress and the confidence of model prediction are improved; the method further utilizes the feature information of the user to be interfered to predict operation to obtain an interference gain score, screens target interference users according to the interference gain score, accurately screens target interference users sensitive to interference operation reaction, avoids the situation that a natural conversion client is judged to be a target interference client, and reduces interference cost.
According to one embodiment of the present application, there is provided a storage medium storing at least one executable instruction for performing the intervention user screening method based on the differential intervention response model in any of the method embodiments described above.
Based on such understanding, the technical solution of the present application may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (may be a CD-ROM, a U-disk, a mobile hard disk, etc.), and includes several instructions for causing a computer device (may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective implementation scenario of the present application.
Fig. 5 is a schematic structural diagram of a computer device according to an embodiment of the present application, and the specific embodiment of the present application is not limited to the specific implementation of the computer device.
As shown in fig. 5, the computer device may include: a processor (processor) 302, a communication interface (Communications Interface) 304, a memory (memory) 306, and a communication bus 308.
Wherein: processor 302, communication interface 304, and memory 306 perform communication with each other via communication bus 308.
A communication interface 304 for communicating with network elements of other devices, such as clients or other servers.
Processor 302 is configured to execute program 310 and may specifically perform the relevant steps in the above-described embodiments of the intervention user screening method based on the differential intervention response model.
In particular, program 310 may include program code including computer-operating instructions.
The processor 302 may be a central processing unit CPU, or an Application-specific integrated Circuit ASIC (Application SPECIFIC INTEGRATED Circuit), or one or more integrated circuits configured to implement embodiments of the present application. The one or more processors included in the computer device may be the same type of processor, such as one or more CPUs; but may also be different types of processors such as one or more CPUs and one or more ASICs.
Memory 306 for storing programs 310. Memory 306 may comprise high-speed RAM memory or may also include non-volatile memory (non-volatile memory), such as at least one disk memory.
Program 310 may be specifically operable to cause processor 302 to:
Acquiring characteristic information of a plurality of groups of users to be intervened;
performing prediction processing on multiple groups of characteristic information based on a differential intervention response model which is trained by the model, obtaining multiple groups of intervention gain scores, wherein the differential intervention response model is constructed based on a multi-task learning neural network model, and model training is completed by utilizing training samples obtained by matching a resampling algorithm based on a tendency score;
and sequencing the intervention gain scores of a plurality of groups, selecting users to be intervened corresponding to the intervention gain scores of a preset number as target intervention users, and executing intervention operation on the target intervention users.
The storage medium may also include an operating system, a network communication module. The operating system is a program that manages the above-described physical device hardware and software resources based on the intervention user screening of the differential intervention response model, supporting the execution of information handling programs and other software and/or programs. The network communication module is used for realizing communication among all components in the storage medium and communication with other hardware and software in the information processing entity equipment.
In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different manner from other embodiments, so that the same or similar parts between the embodiments are mutually referred to. For system embodiments, the description is relatively simple as it essentially corresponds to method embodiments, and reference should be made to the description of method embodiments for relevant points.
The method and system of the present application may be implemented in a number of ways. For example, the methods and systems of the present application may be implemented by software, hardware, firmware, or any combination of software, hardware, firmware. The above-described sequence of steps for the method is for illustration only, and the steps of the method of the present application are not limited to the sequence specifically described above unless specifically stated otherwise. Furthermore, in some embodiments, the present application may also be embodied as programs recorded in a recording medium, the programs including machine-readable instructions for implementing the methods according to the present application. Thus, the present application also covers a recording medium storing a program for executing the method according to the present application.
It will be appreciated by those skilled in the art that the modules or steps of the application described above may be implemented in a general purpose computing device, they may be concentrated on a single computing device, or distributed across a network of computing devices, they may alternatively be implemented in program code executable by computing devices, so that they may be stored in a memory device for execution by computing devices, and in some cases, the steps shown or described may be performed in a different order than that shown or described, or they may be separately fabricated into individual integrated circuit modules, or multiple modules or steps within them may be fabricated into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.
The above description is only of the preferred embodiments of the present application and is not intended to limit the present application, but various modifications and variations can be made to the present application by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.
Claims (7)
1. An intervention user screening method based on a differential intervention response model is characterized by comprising the following steps:
Acquiring characteristic information of a plurality of groups of users to be intervened;
performing prediction processing on multiple groups of characteristic information based on a differential intervention response model which is trained by the model, obtaining multiple groups of intervention gain scores, wherein the differential intervention response model is constructed based on a multi-task learning neural network model, and model training is completed by utilizing training samples obtained by matching a resampling algorithm based on a tendency score;
Sequencing a plurality of groups of intervention gain scores, selecting users to be intervened corresponding to a preset number of the intervention gain scores as target intervention users, and executing intervention operation on the target intervention users;
before the differential intervention response model based on the completed model training predicts multiple sets of the characteristic information, the method further comprises:
Constructing a multitask learning neural network model;
Initializing the multi-task learning neural network model to obtain an initial differential intervention response model;
Calculating conversion probability prediction loss parameters of a first training sample based on a loss function, and updating the initial differential intervention response model by combining a back propagation algorithm and a random gradient descent algorithm, wherein the first training sample is obtained based on a tendency score matching resampling algorithm;
If the conversion probability prediction loss parameters meet preset standards, model training is completed, and a differential intervention response model for completing model training is obtained;
before the calculating the conversion probability prediction loss parameter of the first training sample based on the loss function, the method further includes:
Respectively acquiring characteristic information of the intervention group users and characteristic information of the comparison group users;
Screening intervention characteristic information from the characteristic information of the intervention group user and the characteristic information of the control group user based on a correlation analysis algorithm to obtain a second training sample, wherein the second training sample is used for representing training samples for training a two-classification logistic regression model;
Predicting a third training sample based on a two-class logistic regression model which is trained by the model to obtain a plurality of intervention group user tendency scores and a plurality of control group user tendency scores, wherein the third training sample is used for representing the characteristic information of users except the second training sample;
the method further comprises the steps of after predicting a third training sample based on the two-class logistic regression model with the completed model training to obtain a plurality of intervention group user tendency scores and a plurality of control group user tendency scores:
Sorting the user trend scores of the intervention group to obtain a user trend score sequence of the intervention group, and grouping the user trend score sequence of the intervention group by combining an equal frequency bin algorithm to obtain a plurality of trend score intervals of the intervention group;
screening the characteristic information of the control group user matched with each intervention group trend score interval from the characteristic information of the control group user;
And combining the characteristic information of the intervention group users and the characteristic information of the control group users in the same trend score interval according to a preset proportion to generate a first training sample.
2. The method according to claim 1, characterized in that said performing an intervention operation on said target intervention user, in particular comprises:
Pre-constructing a mapping relation between the intervention operation category and the characteristic information of the user;
determining an intervention operation category matched with the characteristic information of the target intervention user based on the mapping relation;
And calling an intervention operation thread matched with the intervention operation category, and executing intervention operation on the target intervention user based on the intervention operation thread.
3. The method of claim 2, wherein after the performing an intervention operation on the target intervention user based on the intervention operation thread, the method further comprises:
and periodically acquiring intervention operation feedback information according to a preset time interval to serve as a basis for updating the intervention operation category.
4. The method of claim 1, wherein prior to the obtaining the characteristic information of the plurality of sets of users to be intervened, the method further comprises:
Acquiring communication addresses of all users in a database;
sending a registration invitation message of a target item to the communication address;
and receiving registration return information of the full-quantity users, and storing the registration return information in a user database of the target item to serve as characteristic information of the users to be intervened.
5. An intervention user screening device based on a differential intervention response model, comprising:
The first acquisition module is used for acquiring characteristic information of a plurality of groups of users to be interfered;
The first prediction module is used for performing prediction processing on a plurality of groups of characteristic information based on a differential intervention response model which is trained by the model, so as to obtain a plurality of groups of intervention gain scores, wherein the differential intervention response model is constructed based on a multi-task learning neural network model, and training samples obtained by matching a resampling algorithm based on a tendency score are used for completing model training;
The selecting module is used for sorting the intervention gain scores of a plurality of groups, selecting users to be intervened corresponding to the intervention gain scores of a preset number as target intervention users, and executing intervention operation on the target intervention users;
before the first prediction module, the apparatus further comprises:
The construction module is used for constructing a multi-task learning neural network model;
The initialization module is used for initializing the multi-task learning neural network model to obtain an initial differential intervention response model;
The updating module is used for calculating conversion probability prediction loss parameters of a first training sample based on a loss function, updating the initial differential intervention response model by combining a back propagation algorithm and a random gradient descent algorithm, wherein the first training sample is obtained by matching a resampling algorithm based on a tendency score;
The updating module is further used for completing model training if the conversion probability prediction loss parameter meets a preset standard, and obtaining a differential intervention response model for completing model training;
before the updating module, the apparatus further includes:
the second acquisition module is used for respectively acquiring the characteristic information of the intervention group user and the characteristic information of the comparison group user;
The first screening module is used for screening the intervention characteristic information from the characteristic information of the intervention group user and the characteristic information of the control group user based on a correlation analysis algorithm to obtain a second training sample, and the second training sample is used for representing a training sample for training a classification logistic regression model;
The second prediction module is used for predicting a third training sample based on a two-class logistic regression model which is trained by the completed model to obtain a plurality of intervention group user tendency scores and a plurality of comparison group user tendency scores, and the third training sample is used for representing the characteristic information of the users except the second training sample;
After the second prediction module, the apparatus further comprises:
The grouping module is used for sequencing the plurality of intervention group user trend scores to obtain an intervention group user trend score sequence, and grouping the intervention group user trend score sequence by combining an equal frequency bin algorithm to obtain a plurality of intervention group trend score intervals;
The second screening module is used for screening the characteristic information of the comparison group users matched with each intervention group trend score interval from the characteristic information of the comparison group users;
and the merging module is used for merging the characteristic information of the intervention group users and the characteristic information of the comparison group users in the same trend score interval according to a preset proportion to generate a first training sample.
6. A storage medium having stored therein at least one executable instruction for causing a processor to perform operations corresponding to the intervention user screening method based on a differential intervention response model as in any of claims 1-4.
7. A computer device, comprising: the device comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete communication with each other through the communication bus;
the memory is configured to store at least one executable instruction that causes the processor to perform operations corresponding to the intervention user screening method based on the differential intervention response model as set forth in any one of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210279004.2A CN114663143B (en) | 2022-03-21 | 2022-03-21 | Intervention user screening method and device based on differential intervention response model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210279004.2A CN114663143B (en) | 2022-03-21 | 2022-03-21 | Intervention user screening method and device based on differential intervention response model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114663143A CN114663143A (en) | 2022-06-24 |
| CN114663143B true CN114663143B (en) | 2024-06-28 |
Family
ID=82031083
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210279004.2A Active CN114663143B (en) | 2022-03-21 | 2022-03-21 | Intervention user screening method and device based on differential intervention response model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114663143B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116805253B (en) * | 2023-08-18 | 2023-11-24 | 腾讯科技（深圳）有限公司 | Intervention gain prediction method, device, storage medium and computer equipment |
| CN116797053B (en) * | 2023-08-25 | 2023-11-10 | 深圳普菲特信息科技股份有限公司 | Chemical production data analysis method, system and medium based on neural network |
Family Cites Families (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6928398B1 (en) * | 2000-11-09 | 2005-08-09 | Spss, Inc. | System and method for building a time series model |
| CN101699432B (en) * | 2009-11-13 | 2015-01-21 | 黑龙江工程学院 | Ordering strategy-based information filtering system |
| US9483954B2 (en) * | 2013-03-12 | 2016-11-01 | Pearson Education, Inc. | Educational network based intervention |
| US20140289163A1 (en) * | 2013-03-15 | 2014-09-25 | New Frontier Advisors, LLP | Method and Computer Program for Minimizing Trading Costs Subject to a Probability Criterion of Optimality Acceptability |
| US9947028B1 (en) * | 2014-02-27 | 2018-04-17 | Intuit Inc. | System and method for increasing online conversion rate of potential users |
| US20200388360A1 (en) * | 2014-12-10 | 2020-12-10 | Koninklijke Philips N.V. | Methods and systems for using artificial neural networks to generate recommendations for integrated medical and social services |
| US20180150883A1 (en) * | 2016-11-30 | 2018-05-31 | Facebook, Inc. | Content selection for incremental user response likelihood |
| US10423991B1 (en) * | 2016-11-30 | 2019-09-24 | Uber Technologies, Inc. | Implementing and optimizing safety interventions |
| US11188840B1 (en) * | 2017-12-29 | 2021-11-30 | Intuit, Inc. | Machine-learning models to facilitate user retention for software applications |
| US11682474B2 (en) * | 2018-12-12 | 2023-06-20 | International Business Machines Corporation | Enhanced user screening for sensitive services |
| CN109858970B (en) * | 2019-02-02 | 2021-07-02 | 中国银行股份有限公司 | User behavior prediction method, device and storage medium |
| CN110569427B (en) * | 2019-08-07 | 2022-03-01 | 智者四海（北京）技术有限公司 | Multi-target sequencing model training and user behavior prediction method and device |
| CN110995487B (en) * | 2019-12-03 | 2022-04-22 | 深圳市物语智联科技有限公司 | Multi-service quality prediction method and device, computer equipment and readable storage medium |
| CN113469205B (en) * | 2020-03-31 | 2023-01-17 | 阿里巴巴集团控股有限公司 | Data processing method and system, network model and training method thereof, and electronic device |
| CN112100442B (en) * | 2020-11-13 | 2021-02-26 | 腾讯科技（深圳）有限公司 | User tendency recognition method, device, equipment and storage medium |
| CN112446541A (en) * | 2020-11-26 | 2021-03-05 | 上海浦东发展银行股份有限公司 | Fusion classification model establishing method, marketing conversion rate gain prediction method and system |
| CN112527958A (en) * | 2020-12-11 | 2021-03-19 | 平安科技（深圳）有限公司 | User behavior tendency identification method, device, equipment and storage medium |
| CN112529319A (en) * | 2020-12-18 | 2021-03-19 | 平安银行股份有限公司 | Grading method and device based on multi-dimensional features, computer equipment and storage medium |
| CN113763019A (en) * | 2021-01-28 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | User information management method and device |
| CN113434761B (en) * | 2021-06-25 | 2024-02-02 | 平安科技（深圳）有限公司 | Recommendation model training method, device, computer equipment and storage medium |
| CN113704599A (en) * | 2021-07-14 | 2021-11-26 | 大箴(杭州)科技有限公司 | Marketing conversion user prediction method and device and computer equipment |
| CN113641896A (en) * | 2021-07-23 | 2021-11-12 | 北京三快在线科技有限公司 | Model training and recommendation probability prediction method and device |
| CN114139724A (en) * | 2021-11-30 | 2022-03-04 | 支付宝(杭州)信息技术有限公司 | Method and device for training gain model |
-
2022
- 2022-03-21 CN CN202210279004.2A patent/CN114663143B/en active Active
Non-Patent Citations (2)
| Title |
|---|
| 基于因果机制的用户行为研究及应用;郑佳碧;中国博士学位论文全文数据库 信息科技辑;20210815(第8期);I140-16 * |
| 基于隐式反馈数据的用户行为分析及购买预测;卞天宇;中国优秀硕士学位论文全文数据库 信息科技辑;20210215(第2期);I138-3091 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114663143A (en) | 2022-06-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114663143B (en) | Intervention user screening method and device based on differential intervention response model | |
| US20230289828A1 (en) | Data processing method, computer device, and readable storage medium | |
| CN108171280A (en) | A kind of grader construction method and the method for prediction classification | |
| CN111401914A (en) | Risk assessment model training and risk assessment method and device | |
| Christidis et al. | Enabling serverless deployment of large-scale ai workloads | |
| CN113159273B (en) | Neural network training method and related equipment | |
| CN111210072B (en) | Prediction model training and user resource limit determining method and device | |
| CN112328646B (en) | Multitask course recommendation method and device, computer equipment and storage medium | |
| CN112887371B (en) | Edge calculation method and device, computer equipment and storage medium | |
| CN113672375A (en) | Resource allocation prediction method, device, equipment and storage medium | |
| CN117574263A (en) | Sample data generation method, device, equipment and readable storage medium | |
| CN115510948A (en) | Block chain fishing detection method based on robust graph classification | |
| CN108196844A (en) | Information processing method, device, electronic equipment and storage medium | |
| CN113159926A (en) | Loan transaction repayment date determination method and device | |
| CN112036641A (en) | Retention prediction method, device, computer equipment and medium based on artificial intelligence | |
| CN113779116A (en) | Object sorting method, related equipment and medium | |
| CN113138977A (en) | Transaction conversion analysis method, device, equipment and storage medium | |
| CN112365189A (en) | Case distribution method and device | |
| CN112036418A (en) | Method and device for extracting user features | |
| CN110727705A (en) | Information recommendation method and device, electronic equipment and computer-readable storage medium | |
| CN117390455B (en) | Data processing method and device, storage medium and electronic equipment | |
| CN114926267B (en) | Risk data processing method, related equipment and medium | |
| CN113298448B (en) | Lease index analysis method and system based on Internet and cloud platform | |
| CN115099317A (en) | User value classification method and device based on double-layer sub-classification model | |
| CN117952547A (en) | Intelligent examination and approval method and device for express expense based on machine learning |
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 |