CN116304652A

CN116304652A - Data heterogeneous-based joint learning model acquisition method and device

Info

Publication number: CN116304652A
Application number: CN202111511525.8A
Authority: CN
Inventors: 刘嘉; 李增祥
Original assignee: Xinzhi I Lai Network Technology Co ltd
Current assignee: Xinzhi I Lai Network Technology Co ltd
Priority date: 2021-12-06
Filing date: 2021-12-06
Publication date: 2023-06-23

Abstract

The invention relates to the technical field of computers, and provides a method and a device for acquiring a joint learning model based on data isomerism. The method comprises the following steps: constructing test data distribution based on the requirement of the target task on the data distribution; selecting similar data from the other participants that is similar to the target training data based on the target training data and the test data distribution provided by the participants that put forward the demand; training a joint learning model by using target training data and similar data to obtain a target joint learning model; and determining the participants with data distribution meeting the preset requirements as target demanding parties, and providing the target joint learning model for the target demanding parties. The invention ensures that the data for training the model is consistent with the data distribution of the target task, so that the obtained joint learning model has better application effect when being applied to the target task of the demander, the performance of the joint learning model is improved, and the generalization capability of the model is further improved.

Description

Data heterogeneous-based joint learning model acquisition method and device

Technical Field

The invention relates to the technical field of computers, in particular to a method and a device for acquiring a joint learning model based on data isomerism.

Background

Joint learning can effectively help multiple participants perform data usage and machine learning modeling under the requirements of user privacy protection, data security and government regulations. In a joint learning scenario, the model demander may have a small amount of data and may need to train the joint learning model together with other data owners to ensure that the obtained model can be applied to the target task. In this case, the data heterogeneity in the training process of the joint learning model is very challenging. On the one hand, the data of the demand party and the data of other data owners may not be independently and uniformly distributed, and the performance of model inference is reduced due to joint learning training based on all the data of the other data owners; on the other hand, the demand party itself has an insufficient amount of data, so that it may not be independently and co-distributed with the data encountered in the target task, and a model with good performance may not be obtained based on the existing data training model.

Disclosure of Invention

In view of the above, the embodiments of the present invention provide a method, an apparatus, an electronic device, and a computer readable storage medium for obtaining a joint learning model based on data isomerization, so as to solve the problem in the prior art that it is difficult to obtain a model with good performance when data isomerization is performed due to insufficient data volume of a demand party.

In a first aspect of the embodiment of the present invention, a method for acquiring a joint learning model based on data isomerism is provided, including:

constructing test data distribution based on the requirement of the target task on the data distribution;

selecting similar data from other participants that is similar to the target training data based on the target training data and the test data distribution provided by the requesting participant;

training a joint learning model by adopting the target training data and the similar data to acquire a target joint learning model;

and determining the participants with data distribution meeting the preset requirements as target demanding parties, and providing the target joint learning model for the target demanding parties.

In a second aspect of the embodiment of the present invention, there is provided a joint learning model acquisition apparatus based on data isomerization, including:

the test data distribution construction module is configured to construct test data distribution based on the requirement of the target task on data distribution;

a similar data acquisition module configured to select similar data from other participants that is similar to the target training data based on the target training data provided by the requesting participant and the test data distribution;

a training module configured to train a joint learning model using the target training data and the similarity data to obtain a target joint learning model;

and the distribution module is configured to determine that the participants with data distribution meeting the preset requirements are target demander and provide the target joint learning model for the target demander.

In a third aspect of the embodiments of the present invention, there is provided an electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the above method when the computer program is executed.

In a fourth aspect of the embodiments of the present invention, there is provided a computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of the above method.

Compared with the prior art, the embodiment of the invention has the beneficial effects that: according to the method and the device for training the model, based on the requirements of the target task on data distribution, test data distribution is built, the target training data provided by the party who puts forward the requirements is combined, similar data similar to the target training data is selected from other parties to carry out data isomerism, the target training data and the similar data are adopted to train the joint learning model, so that the target joint learning model is obtained, finally the target joint learning model is provided for the target requiring party, the data for training the model are consistent with the data distribution of the target task, the joint learning model has better application effect when the joint learning model is applied to the target task of the requiring party, the performance of the joint learning model is improved, and the popularization capability of the model is further improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic diagram of a joint learning architecture according to an embodiment of the present invention;

FIG. 2 is a flowchart of a method for acquiring a joint learning model based on data isomerism according to an embodiment of the invention;

FIG. 3 is a schematic diagram of a joint learning model acquisition device based on data isomerism according to an embodiment of the invention;

fig. 4 is a schematic diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

In the following description, for purposes of explanation and not limitation, specific details are set forth such as the particular system architecture, techniques, etc., in order to provide a thorough understanding of the embodiments of the present invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.

The joint learning refers to comprehensively utilizing a plurality of AI (Artificial Intelligence ) technologies on the premise of ensuring data safety and user privacy, jointly excavating data value by combining multiparty cooperation, and promoting new intelligent business states and modes based on joint modeling. The joint learning has at least the following characteristics:

(1) The participating nodes control the weak centralized joint training mode of the own data, so that the data privacy safety in the co-creation intelligent process is ensured.

(2) Under different application scenes, a plurality of model aggregation optimization strategies are established by utilizing screening and/or combination of an AI algorithm and privacy protection calculation so as to obtain a high-level and high-quality model.

(3) On the premise of ensuring data safety and user privacy, a method for improving the efficiency of the joint learning engine is obtained based on a plurality of model aggregation optimization strategies, wherein the efficiency method can be used for improving the overall efficiency of the joint learning engine by solving the problems of information interaction, intelligent perception, exception handling mechanisms and the like under a large-scale cross-domain network with parallel computing architecture.

(4) The requirements of multiparty users in all scenes are acquired, the real contribution degree of all joint participants is determined and reasonably evaluated through a mutual trust mechanism, and distribution excitation is carried out.

Based on the mode, AI technical ecology based on joint learning can be established, the industry data value is fully exerted, and the scene of the vertical field is promoted to fall to the ground.

A method and apparatus for acquiring a joint learning model based on data isomerization according to embodiments of the present disclosure will be described in detail with reference to the accompanying drawings.

Fig. 1 is a schematic diagram of a joint learning architecture according to an embodiment of the present disclosure. As shown in fig. 1, the architecture of joint learning may include a server (central node) 101, as well as

participants

102, 103, and 104.

In the joint learning process, a basic model may be established by the server 101, and the server 101 transmits the model to the

participants

102, 103, and 104 with which a communication connection is established. The basic model may also be uploaded to the server 101 after any party has established, and the server 101 sends the model to the other parties with whom it has established a communication connection. The

participants

102, 103 and 104 construct a model according to the downloaded basic structure and model parameters, perform model training using local data, obtain updated model parameters, and encrypt and upload the updated model parameters to the server 101. Server 101 aggregates the model parameters sent by

participants

102, 103, and 104 to obtain global model parameters, and transmits the global model parameters back to

participants

102, 103, and 104. Participant 102, participant 103 and participant 104 iterate the respective models according to the received global model parameters until the models eventually converge, thereby enabling training of the models. In the joint learning process, the data uploaded by the

participants

102, 103 and 104 are model parameters, local data is not uploaded to the server 101, and all the participants can share final model parameters, so that common modeling can be realized on the basis of ensuring data privacy. It should be noted that the number of participants is not limited to three as described above, but may be set as needed, and the embodiment of the present disclosure is not limited thereto.

Fig. 2 is a flowchart of a method for acquiring a joint learning model based on data isomerism according to an embodiment of the present invention. The data heterogeneous based joint learning model acquisition method of fig. 2 may be performed by the server 101 in fig. 1. As shown in fig. 2, the data heterogeneous-based joint learning model acquisition method includes:

s201, based on the requirement of the target task on the data distribution, the test data distribution is constructed.

The target task may be a target task of a party who makes a request, or may be a target task of another user, which is not limited herein. For different target tasks, the data distribution may be different, so that in order to ensure that the obtained joint learning model can be effectively applied to the target tasks, it is required to ensure that the training data distribution in the model training process is consistent with the data distribution requirement of the target tasks. Therefore, it is necessary to construct a test data distribution based on the requirement of the target task for the data distribution first, the test data distribution being consistent with the requirement of the data distribution of the target task.

In this embodiment, the construction of the test data distribution may include: determining data distribution characteristics of the target task based on the historical experience information; based on the data distribution characteristics, test data distribution of the target task is constructed. Wherein the historical experience information may be based on artificial experience knowledge summarized for the target task. Because the target task is a task in future application, the test data distribution of the target task cannot be directly obtained, and the test data distribution of the target task can be predicted according to historical experience information, so that the test data distribution can be constructed based on the historical experience information, and the accuracy of the test data distribution is ensured.

S202, based on the target training data and the test data distribution provided by the party making the request, similar data similar to the target training data is selected from the other parties.

In the process of model training, a plurality of participants can participate in the model training, wherein the participant who puts forward the model requirement is a target participant, initiates the model training, and other participants respond to the requirements of the target participant and add into the model training process. During model training, each participant may provide training data. Considering that the data provided by other participants may contain data that is not relevant to model training, which is quite different from the test data distribution, if it is incorporated into the model training process, the effect of model training is not improved, but the accuracy of the obtained model is reduced. Thus, it is necessary to select similar data similar to the target training data from each other party for model training based on the target training data and the test data distribution situation provided by the target party.

In this embodiment, based on the target training data and the test data distribution provided by the party making the request, selecting similar data similar to the target training data from other parties may specifically include:

the training data of the party (target party) who puts forward the demand is obtained to constitute the training data set, and the training data distribution in the training data set is obtained simultaneously, and the quantity of training data can be set up as required, and the more training data that the target party provided, the higher the training precision of model.

And sequentially taking each piece of training data in the training data distribution as target training data, and determining the target quantity of similar data according to the test data distribution. The distribution of the training data may be different from the distribution of the test data, so that the data distribution of the training data is required to be consistent with the distribution of the test data in order to ensure that the trained model is consistent with the required model. The method for determining the target quantity comprises the following steps:

n＝P-test(i)/P-train(i)

wherein, P-train represents the training data distribution, P-train (i) is the i-th training data distribution in the training data distribution; p-test represents the test data distribution, and P-test (i) is the distribution corresponding to the ith training data in the test data distribution; n represents the target number, n is an integer taken downwards, and the range of values of i can be 1,2,3, …, n and the like.

For each piece of training data, the corresponding situation in the test data distribution is different, and thus the corresponding quantity is different, for example, for the first piece of training data, the quantity of similar data in the case of meeting the test data distribution according to the test data distribution situation is 2 pieces; for the second piece of training data, the number of similar data should be 10 in case the test data distribution is satisfied, and so on. By sequentially determining each piece of training data as target training data, the target number of corresponding similar data can be obtained, thereby facilitating subsequent determination of similar data.

And confirming the data which satisfies the similarity requirement with the target training data in the other participants as similar data, wherein the number of the similar data is consistent with the target number. After the target number of the similar data is determined, selecting the data meeting the similarity requirement from other participants as the similar data according to the target number to participate in subsequent model training. Specifically, the data in other participants may be ranked according to the similarity between the data and the target training data, and the target number of data arranged in front may be sequentially used as the similar data of the target training data, where the similar data may originate from the same other participant or may originate from different other participants.

In consideration of the safety of the data, when the similarity between other participants and the target training data is determined, the data can be subjected to privacy protection based on data encryption modes such as local sensitive hash and the like, so that the data leakage is avoided. Taking a local sensitive hash algorithm (Locality Sensitive Hashing, LSH) as an example, a party which puts forward a requirement encrypts target training data to obtain a first hash code, and uploads the first hash code to a server of a joint learning architecture; the other participants encrypt the data respectively to obtain a second hash code, and upload the second hash code to a server of the joint learning architecture; and obtaining the similarity of each second hash code and the first hash code, sequencing according to the similarity, sequentially determining the target number of second hash codes arranged in front as target hash codes, and determining the data corresponding to the target hash codes as similar data. Of course, the data may be privacy-protected by other encryption methods, and is not limited to the above case.

S203, training the joint learning model by using the target training data and the similar data to acquire the target joint learning model.

On the one hand, constructing a training data set from data related to training data distribution of the participants who put forward the requirements; on the other hand, the similar data is duplicated for corresponding times according to the selected times of each similar data in each other participant, and corresponding quantity of similar data is obtained, so that a training data set of each other participant is obtained; the joint learning model is then trained according to the training data set of the party that placed the demand and the training data set of each of the other parties to obtain a target joint learning model. And the data of each participant does not need to be uploaded in the training process, but the obtained model parameters are uploaded to a server after the model is trained locally and respectively, and the server aggregates the parameters and then transmits the parameters to each participant again so as to carry out the next round of model training. The process of model training may be performed in multiple rounds until a target joint learning model that meets the requirements is obtained.

S204, determining the party with the data distribution meeting the preset requirement as a target demand party, and providing the target joint learning model for the target demand party.

In this embodiment, the party making the demand may be the target party, or may be another party participating in training, or may be another party, which is not limited herein. In order to ensure that the desiring party can obtain the ideal effect when using the target joint learning model, the desiring party needs to be screened.

For example, the demand party may be screened based on the test data distribution of the target joint learning model, where the data distribution of the participant is obtained according to the model providing demand, and the participant is not limited to the participant in the model training process; judging whether the consistency of the data distribution of the participator and the test data distribution is greater than a preset threshold value; if the consistency of the data distribution of the participant and the test data distribution is greater than a preset threshold, the data distribution of the participant is consistent with the data distribution of the target joint learning model, and the target joint learning model can have good application effect when being applied to the target task of the participant, so that the participant is determined to be the target demand side, and the target joint learning model is provided for the target demand side; if the consistency of the data distribution of the participant and the test data distribution is smaller than the preset threshold, the data distribution of the participant is inconsistent with the data distribution of the target joint learning model, and the target joint learning model cannot be applied to the target task of the participant, so that the participant is not the target demand side.

For another example, since the target training model is obtained after training based on training data of the target participant and is also an application requirement of the target participant, the demander can be screened based on the training data distribution of the participant (target participant) who puts out the requirement. At this time, according to the model providing requirement, the data distribution of the participants is obtained, and the participants are not limited to the participants in the model training process; judging whether the consistency of the data distribution of the participant and the training data distribution of the target participant is greater than a preset threshold value; if the consistency of the data distribution of the participant and the training data distribution is greater than a preset threshold, the data distribution of the participant is consistent with the training data distribution of the target participant, and the target joint learning model can have good application effect when being applied to the target task of the participant, so that the participant is determined to be the target demand side, and the target joint learning model is provided for the target demand side; if the consistency of the data distribution of the participant and the training data distribution is smaller than the preset threshold value, the data distribution of the participant is inconsistent with the training data distribution of the target participant, and the target joint learning model cannot be applied to the target task of the participant, so that the participant is not the target demand side.

The technical scheme provided by the embodiment of the disclosure at least comprises the following beneficial effects: according to the method and the device, firstly, the requirement of a target task on data distribution is determined based on historical experience information, test data distribution is constructed, target training data provided by a party who proposes the requirement is combined, similar data similar to the target training data is selected from other parties to carry out data isomerism, a target combined learning model is trained by the target training data and the similar data so as to obtain the target combined learning model, and finally, the target combined learning model is provided for a target requiring party.

Any combination of the above optional solutions may be adopted to form an optional embodiment of the present application, which is not described herein in detail.

The following are examples of the apparatus of the present invention that may be used to perform the method embodiments of the present invention. For details not disclosed in the embodiments of the apparatus of the present invention, please refer to the embodiments of the method of the present invention.

Fig. 3 is a schematic diagram of a joint learning model acquisition device based on data isomerism according to an embodiment of the invention. As shown in fig. 3, the data heterogeneous based joint learning model acquisition apparatus includes a test data distribution construction module 301, a similar data acquisition module 302, a training module 303, and an allocation module 304. Wherein the test data distribution construction module 301 is configured to construct a test data distribution based on the requirement of the target task for the data distribution; the similar data acquisition module 302 is configured to select similar data from other participants that is similar to the target training data based on the target training data and the test data distribution provided by the requesting participant; the training module 303 is configured to train the joint learning model using the target training data and the similar data to obtain a target joint learning model; the assignment module 304 is configured to determine that the participants whose data distribution meets the preset requirements are target requesters and to provide the target joint learning model to the target requesters.

Further, the test data distribution construction module 301 is specifically configured to: determining data distribution characteristics of the target task based on the historical experience information; based on the data distribution characteristics, test data distribution of the target task is constructed.

Further, the similar data acquisition module 302 is specifically configured to: acquiring training data distribution of a party who puts forward the requirement; taking each piece of training data in the training data distribution as target training data in sequence, and determining the target quantity of similar data according to the test data distribution; and confirming the data which satisfies the similarity requirement with the target training data in the other participants as similar data, wherein the number of the similar data is consistent with the target number.

Further, the training module 303 is specifically configured to: taking data related to the training data distribution of the participants who put forward the requirements as training data to construct a training data set of the participants who put forward the requirements; copying the similar data according to the selected times of each similar data in each other participant to obtain a training data set of each other participant; and training the joint learning model according to the training data set of the party which puts forward the requirement and the training data set of each other party so as to acquire a target joint learning model.

In some embodiments, the assignment module 304 is specifically configured to: obtaining data distribution of the participants according to the model providing requirements; judging whether the consistency of the data distribution of the participator and the test data distribution is greater than a preset threshold value; if the consistency of the data distribution of the participant and the test data distribution is greater than a preset threshold, determining that the participant is a target demand side; the target joint learning model is provided to the target demander.

In other embodiments, the distribution module 304 is specifically configured to: obtaining data distribution of the participants according to the model providing requirements; judging whether the consistency of the data distribution of the participants and the training data distribution of the participants who put forward the requirements is larger than a preset threshold value; if the consistency of the data distribution of the participant and the training data distribution is greater than a preset threshold, determining that the participant is a target demand side; the target joint learning model is provided to the target demander.

According to the method and the device for training the data distribution, based on the requirements of the target task on the data distribution, test data distribution is built, the target training data provided by the party who puts forward the requirements is combined, similar data similar to the target training data is selected from other parties to carry out data isomerism, the target training data and the similar data are adopted to train the joint learning model, so that the target joint learning model is obtained, finally, the target joint learning model is provided for the target party, the data for training the model is ensured to be consistent with the data distribution of the target task through adding historical experience information into the data isomerism process, so that the obtained joint learning model has better application effect when being applied to the target task of the party, the performance of the joint learning model is improved, and the popularization capability of the model is further improved.

It should be understood that the sequence number of each step in the foregoing embodiment does not mean that the execution sequence of each process should be determined by the function and the internal logic, and should not limit the implementation process of the embodiment of the present invention.

Fig. 4 is a schematic diagram of an electronic device 4 according to an embodiment of the present invention. As shown in fig. 4, the electronic apparatus 4 of this embodiment includes: a processor 401, a memory 402 and a computer program 403 stored in the memory 402 and executable on the processor 401. The steps of the various method embodiments described above are implemented by processor 401 when executing computer program 403. Alternatively, the processor 401, when executing the computer program 403, performs the functions of the modules/units in the above-described apparatus embodiments.

Illustratively, the computer program 403 may be partitioned into one or more modules/units, which are stored in the memory 402 and executed by the processor 401 to complete the present invention. One or more of the modules/units may be a series of computer program instruction segments capable of performing a specific function for describing the execution of the computer program 403 in the electronic device 4.

The electronic device 4 may be a desktop computer, a notebook computer, a palm computer, a cloud server, or the like. The electronic device 4 may include, but is not limited to, a processor 401 and a memory 402. It will be appreciated by those skilled in the art that fig. 4 is merely an example of the electronic device 4 and is not meant to be limiting of the electronic device 4, and may include more or fewer components than shown, or may combine certain components, or different components, e.g., the electronic device may also include an input-output device, a network access device, a bus, etc.

The processor 401 may be a central processing unit (Central Processing Unit, CPU) or other general purpose processor, digital signal processor (Digital Signal Processor, DSP), application specific integrated circuit (Application Specific Integrated Circuit, ASIC), field programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The memory 402 may be an internal storage unit of the electronic device 4, for example, a hard disk or a memory of the electronic device 4. The memory 402 may also be an external storage device of the electronic device 4, for example, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash Card (Flash Card) or the like, which are provided on the electronic device 4. Further, the memory 402 may also include both internal storage units and external storage devices of the electronic device 4. The memory 402 is used to store computer programs and other programs and data required by the electronic device. The memory 402 may also be used to temporarily store data that has been output or is to be output.

It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working process of the units and modules in the above system may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.

In the foregoing embodiments, the descriptions of the embodiments are emphasized, and in part, not described or illustrated in any particular embodiment, reference is made to the related descriptions of other embodiments.

Those of ordinary skill in the art will appreciate that the elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus/electronic device and method may be implemented in other manners. For example, the apparatus/electronic device embodiments described above are merely illustrative, e.g., the division of modules or elements is merely a logical functional division, and there may be additional divisions of actual implementations, multiple elements or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection via interfaces, devices or units, which may be in electrical, mechanical or other forms.

The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated modules/units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the present invention may implement all or part of the flow of the method of the above embodiment, or may be implemented by a computer program to instruct related hardware, and the computer program may be stored in a computer readable storage medium, where the computer program, when executed by a processor, may implement the steps of each of the method embodiments described above. The computer program may comprise computer program code, which may be in source code form, object code form, executable file or in some intermediate form, etc. The computer readable medium may include: any entity or device capable of carrying computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), an electrical carrier signal, a telecommunications signal, a software distribution medium, and so forth. It should be noted that the content of the computer readable medium can be appropriately increased or decreased according to the requirements of the jurisdiction's jurisdiction and the patent practice, for example, in some jurisdictions, the computer readable medium does not include electrical carrier signals and telecommunication signals according to the jurisdiction and the patent practice.

The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention, and are intended to be included in the scope of the present invention.

Claims

1. The method for acquiring the joint learning model based on the data isomerism is characterized by comprising the following steps of:

2. The method of claim 1, wherein the constructing a test data distribution step based on the demand for data distribution by the target task comprises:

determining data distribution characteristics of the target task based on the historical experience information;

and constructing test data distribution of the target task based on the data distribution characteristics.

3. The method of claim 1, wherein the selecting similar data from other participants that is similar to the target training data based on the target training data and the test data distribution provided by the requesting participant comprises:

acquiring training data distribution of a party who puts forward the requirement;

taking each piece of training data in the training data distribution as target training data in sequence, and determining the target quantity of similar data according to the test data distribution;

and confirming data which satisfies the similarity requirement with the target training data in other participants as similar data, wherein the number of the similar data is consistent with the target number.

4. A method according to claim 3, wherein said identifying data in the other participants that satisfy the similarity requirement for similarity to the target training data as similar data comprises:

the participant who puts forward the demand encrypts the target training data to obtain a first hash code, and uploads the first hash code to a server of a joint learning architecture;

the other participants encrypt the data respectively to obtain a second hash code, and upload the second hash code to a server of the joint learning architecture;

obtaining the similarity of each second hash code and the first hash code, and selecting a target number of second hash codes with the similarity meeting preset requirements as target hash codes;

and confirming the data corresponding to the target hash code as similar data.

5. The method of claim 3, wherein training a joint learning model using the target training data and the similarity data to obtain a target joint learning model comprises:

taking data related to the training data distribution of the participants who put forward the requirements as training data to construct a training data set of the participants who put forward the requirements;

copying the similar data according to the selected times of each similar data in each other participant to obtain a training data set of each other participant;

and training the joint learning model according to the training data set of the party which puts forward the requirement and the training data set of each other party so as to acquire a target joint learning model.

6. The method of claim 1, wherein the determining that the data distribution meets the preset requirement is a target demander and providing the target joint learning model to the target demander comprises:

obtaining data distribution of the participants according to the model providing requirements;

judging whether the consistency of the data distribution of the participator and the test data distribution is greater than a preset threshold value;

if the consistency of the data distribution of the participant and the test data distribution is greater than a preset threshold, determining that the participant is a target demand side;

and providing the target joint learning model to the target demander.

7. The method of claim 1, wherein the determining that the data distribution meets the preset requirement is a target demander and providing the target joint learning model to the target demander comprises:

judging whether the consistency of the data distribution of the participants and the training data distribution of the participants who put forward the requirements is larger than a preset threshold value;

if the consistency of the data distribution of the participant and the training data distribution is greater than a preset threshold, determining that the participant is a target demand side;

and providing the target joint learning model to the target demander.

8. A joint learning model acquisition device based on data heterogeneity, comprising:

9. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any of claims 1 to 7 when the computer program is executed.

10. A computer readable storage medium storing a computer program, characterized in that the computer program when executed by a processor implements the steps of the method according to any one of claims 1 to 7.