Model training method, device, equipment and storage medium

Abstract

The present disclosure provides a model training method, device, equipment and storage medium, which relate to the field of computer technology, and in particular to the field of artificial intelligence technology. The specific implementation scheme is: verifying the permission information of the first object and the original sample data of the second object; when it is determined that the permission information and the original sample data have passed the verification, determining the target sample data according to the original sample data; and using the target sample data to train the deep learning model to obtain the target model.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to the field of artificial intelligence technology. Background Art

[0002] Joint modeling refers to the process of building a data model of an information system based on data from multiple domain systems using formal data modeling techniques. Summary of the invention

[0003] The present invention provides a model training method, apparatus, device and storage medium.

[0004] According to one aspect of the present disclosure, a model training method is provided, including: verifying permission information of a first object and original sample data of a second object; when it is determined that both the permission information and the original sample data have passed the verification, determining target sample data based on the original sample data; and using the target sample data to train a deep learning model to obtain a target model.

[0005] According to another aspect of the present disclosure, a model training device is provided, including: a verification module for verifying the permission information of a first object and the original sample data of a second object; a data processing module for determining target sample data based on the original sample data when it is determined that the permission information and the original sample data have passed the verification; and a model training module for training a deep learning model using the target sample data to obtain a target model.

[0006] Another aspect of the present disclosure provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method shown in an embodiment of the present disclosure.

[0007] According to another aspect of an embodiment of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the method shown in the embodiment of the present disclosure.

[0008] According to another aspect of an embodiment of the present disclosure, a computer program product is provided, including a computer program / instruction, wherein the computer program / instruction implements the steps of the method shown in the embodiment of the present disclosure when executed by a processor.

[0009] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0011] Figure 1 It is a schematic diagram of an application scenario of a model training method, device, electronic device, and storage medium according to an embodiment of the present disclosure;

[0012] Figure 2 A flow chart of a model training method according to an embodiment of the present disclosure is schematically shown;

[0013] Figure 3 A flowchart schematically illustrates a method for configuring permission information of a first object according to an embodiment of the present disclosure;

[0014] Figure 4 A flowchart schematically illustrates a method for acquiring original sample data according to an embodiment of the present disclosure;

[0015] Figure 5 A flow chart of a model training method according to an embodiment of the present disclosure is schematically shown;

[0016] Figure 6 A flow chart of a model training method according to an embodiment of the present disclosure is schematically shown;

[0017] Figure 7 A schematic diagram schematically shows a model training method according to another embodiment of the present disclosure;

[0018] Figure 8 A block diagram schematically shows a model training device according to an embodiment of the present disclosure; and

[0019] Fig. 9 A block diagram schematically illustrates an example electronic device that may be used to implement an embodiment of the present disclosure. DETAILED DESCRIPTION

[0020] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0021] The following will be combined Figure 1 The application scenarios of the model training method and device provided in the present disclosure are described.

[0022] Figure 1Schematic diagram of application scenarios of the model training method, device, electronic device and storage medium according to the embodiment of the present disclosure. It should be noted that Figure 1 What is shown is merely an example of a system architecture to which the embodiments of the present disclosure can be applied, in order to help those skilled in the art understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used in other devices, systems, environments or scenarios.

[0023] like Figure 1 As shown, the system architecture 100 according to this embodiment may include operating platforms 101, 102, a network 103 and a server 104. The network 103 is used to provide a medium for a communication link between the operating platforms 101, 102 and the server 104. The network 103 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0024] Users can use the operating platforms 101 and 102 to interact with the server 104 through the network 103 to receive or send messages, etc.

[0025] The operating platforms 101 and 102 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0026] Server 104 can be a project management server for a deep learning model training project. Server 104 can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services (Virtual Private Server, or VPS for short). Server 104 can also be a server for a distributed system, or a server combined with a blockchain.

[0027] According to an embodiment of the present disclosure, the entrusting party may entrust the entrustee to train the deep learning model on its behalf, that is, to perform joint modeling. To this end, the entrusting party may sign a delegation agreement with the entrustee. According to an embodiment of the present disclosure, the entrusting party may upload the entrusting party's signature information to the server 104 through the operating platform 101, and on the other hand, the entrustee may upload the entrustee's signature information to the server 104 through the operating platform 102. The server 104 may verify the entrusting party's signature information and the entrustee's signature information, and if the verification passes, configure the entrustee with permission information. The permission information is used to indicate that the entrustee has the relevant hardware use permission and software use permission.

[0028] The entrusted party can submit data demand information to the server 104 through the operating platform 102. The server 104 can determine the demand request based on the data demand information and send it to the operating platform 101. The entrusting party can prepare the corresponding original sample data based on the demand request and upload the original sample data to the server 104.

[0029] Before model training, the server 104 verifies the authority information of the entrusted party and the original sample data provided by the entrusting party. When it is determined that both the authority information and the original sample data have passed the verification, the original sample data can be preprocessed to obtain the target sample data. Then, the deep learning model can be trained using the target sample data to obtain the target model.

[0030] According to the model training method of the embodiment of the present disclosure, the work of the entrusting party and the entrusting party can be coordinated to prevent duplication of work, prevent process disruption, and improve the efficiency of training deep learning models.

[0031] It should be understood that Figure 1 The number of operating platforms, networks and servers in the embodiment is only for illustration. Any number of operating platforms, networks and servers may be provided as required.

[0032] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision, disclosure and application of user personal information involved comply with the provisions of relevant laws and regulations, take necessary confidentiality measures, and do not violate public order and good morals.

[0033] In the technical solution of the present disclosure, the user's authorization or consent is obtained before obtaining or collecting the user's personal information.

[0034] Figure 2 A flowchart of a model training method according to an embodiment of the present disclosure is schematically shown.

[0035] like Figure 2 As shown, the model training method 200 includes, in operation S210, verifying the permission information of the first object and the original sample data of the second object.

[0036] Then, in operation S220, when it is determined that both the authority information and the original sample data have passed the verification, the target sample data is determined according to the original sample data.

[0037] In operation S230, a deep learning model is trained using the target sample data to obtain a target model.

[0038] According to an embodiment of the present disclosure, the second object may include, for example, a user of the entrusting party. The first object may include, for example, a user of the entrusting party. The entrusting party may entrust the entrusting party to train the deep learning model on its behalf, that is, to perform joint modeling.

[0039] According to an embodiment of the present disclosure, the permission information of the first object may be used to indicate whether the first object has permission to use corresponding software and hardware. The software and hardware may be software and hardware for training a deep learning model.

[0040] According to an embodiment of the present disclosure, the original sample data may be provided by the second object, for example.

[0041] According to an embodiment of the present disclosure, the permission information of the first object can be checked to determine whether the first object has permission to use the corresponding software and hardware. If it is determined that the first object has permission to use the corresponding software and hardware, it is determined that the permission information of the first object passes the verification. Otherwise, it is determined that the permission information of the first object fails the verification.

[0042] According to an embodiment of the present disclosure, it is possible to verify whether the original sample data contains all the fields required for training. If it is determined that the original sample data contains all the fields required for training, it is determined that the original sample data passes the verification. Otherwise, it is determined that the original sample data fails the verification.

[0043] According to an embodiment of the present disclosure, the original sample data may include multiple data tables. For example, multiple target fields may be extracted from the multiple data tables. Then, based on the multiple target fields, a data wide table is determined as the target sample data. Among them, the target field may be a field corresponding to the sample data required for training the deep learning model.

[0044] According to the embodiments of the present disclosure, the first object can call the corresponding software and hardware to train the deep learning model, and during the training process, the original data samples provided by the second object can be used. In this way, the work of the first object and the second object can be coordinated to prevent duplication of work, prevent process disorder, and improve the efficiency of model building.

[0045] According to another embodiment of the present disclosure, the above method may further include configuring permission information of the first object. Figure 3 The method for configuring the permission information of the first object according to the embodiment of the present disclosure is further described.

[0046] Figure 3 The flowchart of the method for configuring the permission information of the first object according to the embodiment of the present disclosure is schematically shown.

[0047] like Figure 3 As shown, the method 300 for configuring permission information of a first object may include, in operation S310, receiving first signature information from a first operating platform and second signature information from a second operating platform.

[0048] According to an embodiment of the present disclosure, the first operating platform corresponds to the first object, and the second operating platform corresponds to the second object. In this embodiment, the first operating platform may include, for example, an operating platform used by the first object, and the second operating platform may include, for example, an operating platform used by the second object.

[0049] It should be noted that the first object may use the same operating platform each time, or may use a different operating platform each time. The second object may use the same operating platform each time, or may use a different operating platform each time. This disclosure does not specifically limit this.

[0050] According to an embodiment of the present disclosure, a first subject may upload first signature information through a first operating platform. A second subject may upload second signature information through a second operating platform. The first signature information may include, for example, a signature image and / or a seal image of the first subject, and the second signature information may include, for example, a signature image and / or a seal image of the second subject.

[0051] Then, in operation S320, the first signature information and the second signature information are verified.

[0052] According to an embodiment of the present disclosure, the signature image and / or seal image of the first object may be verified to determine the authenticity of the signature in the signature image and / or the seal in the seal image. If any one of the signature or the seal is false, it is determined that the first signature information has not passed the verification. Otherwise, it is determined that the first signature information has passed the verification.

[0053] Similarly, the signature image and / or seal image of the second object can be verified to determine the authenticity of the signature in the signature image and / or the seal in the seal image. If any one of the signature or seal is false, it is determined that the first signature information has not passed the verification. Otherwise, it is determined that the first signature information has passed the verification.

[0054] In operation S330, when both the first signature information and the second signature information pass verification, permission information of the first object is configured to enable the first object to have permission to use the hardware and software corresponding to the model training.

[0055] According to an embodiment of the present disclosure, if both the first signature information and the second signature information pass the verification, it means that the first object and the second object have reached an agreement, and the corresponding permission information can be configured for the first object to open up the usage rights of the hardware and software corresponding to the model training for the first object.

[0056] According to another embodiment of the present disclosure, the above method may further include obtaining original sample data. The original sample data may be determined by the second object according to the data requirements of the first object and uploaded through the second operating platform. Figure 4The method for obtaining original sample data according to an embodiment of the present disclosure is further described.

[0057] Figure 4 The flowchart of the method for acquiring original sample data according to an embodiment of the present disclosure is schematically shown.

[0058] like Figure 4 As shown, the method 400 for obtaining original sample data may include, in operation S410, in response to receiving data demand information from a first operating platform, generating a demand request according to the data demand information.

[0059] According to an embodiment of the present disclosure, the first object may, for example, upload data requirement information through the first operating platform. The data requirement information may, for example, include the data type, format, quantity, etc. required for training the deep learning model.

[0060] In operation S420, the demand request is sent to the second operating platform.

[0061] According to an embodiment of the present disclosure, the first object may, for example, receive a demand request through a second operating platform, and then prepare corresponding data, namely original sample data, according to the demand request, and then upload the original sample data.

[0062] In operation S430, raw sample data is received from the second operating platform.

[0063] According to an embodiment of the present disclosure, after receiving the original sample data from the second operating platform, the original sample data may be stored for subsequent use.

[0064] Figure 5 A flowchart of a model training method according to an embodiment of the present disclosure is schematically shown.

[0065] like Figure 5 As shown, the model training method 500 may include, in operation S510, performing data preprocessing on target sample data to obtain a feature file.

[0066] In operation S520, the feature file is input into a deep learning model to obtain an output result.

[0067] In operation S530, parameters of the deep learning model are adjusted according to the output result.

[0068] According to an embodiment of the present disclosure, data preprocessing may include, for example, any one or more of the following processes: data conversion, feature processing, encoding of text, character string and other types of data, discretization of continuous data, and the like.

[0069] According to an embodiment of the present disclosure, through preprocessing, the target sample data can be converted into a feature file that can be input into a deep learning model.

[0070] According to an embodiment of the present disclosure, during the training process, it can be determined whether the deep learning model meets the predetermined conditions. If the predetermined conditions are met, the training can be stopped, and the trained deep learning model can be determined as the target model. If the predetermined conditions are not met, operations S520 to S530 can be continuously executed in a loop until the predetermined conditions are met. Among them, meeting the predetermined conditions may include, for example, the deep learning model converges or a specific indicator of the deep learning model reaches a predetermined requirement. Among them, the specific indicator and the corresponding predetermined requirement are set according to the needs.

[0071] According to an embodiment of the present disclosure, after the deep learning model completes training, the trained deep learning model can be put online for use by the second object.

[0072] Based on this, Figure 6 A flowchart of a model training method according to an embodiment of the present disclosure is schematically shown.

[0073] like Figure 6 As shown, the model training method 600 may further include, in operation S610, generating an access interface corresponding to the target model.

[0074] In operation S620, target input data is received from the second operating platform through the access interface.

[0075] In operation S630, target output data corresponding to the target input data is determined using the target model.

[0076] In operation S640, the target output data is sent to the second operating platform through the access interface.

[0077] According to an embodiment of the present disclosure, an access interface can be used to access a target model. The access interface can be provided to a second operating platform. The second object can send data to be input into the target model, i.e., target input data, to the access interface through the second operating platform, so as to determine target output data corresponding to the target input data using the target model. After obtaining the target output data, the target output data can be returned to the second operating platform through the access interface.

[0078] According to another embodiment of the present disclosure, the above method may further include performing data exploration on the target sample data to obtain sample distribution information. The target sample data may be adjusted according to the sample distribution information, for example, some sample data may be reduced or some sample data may be increased, so that the sample data better meets actual needs.

[0079] According to another embodiment of the present disclosure, the method may further include performing a model evaluation on the target model to obtain an algorithm index and a business index, and then optimizing the target model when it is determined that at least one of the algorithm index and the business index is less than a corresponding threshold.

[0080] According to an embodiment of the present disclosure, algorithm indicators may include, for example, accuracy, recall, ROC (receiver operating characteristic curve), etc. Business indicators may include, for example, user conversion rate, sales increase rate, etc.

[0081] According to the embodiments of the present disclosure, corresponding thresholds may be set for each algorithm indicator and business indicator according to actual requirements of the first object and the second object.

[0082] According to an embodiment of the present disclosure, the target model may be optimized, for example, by adjusting parameters of the target model, changing the structure of the target model, or replacing it with another type of deep learning model.

[0083] According to another embodiment of the present disclosure, the above method may further include interpreting the target model, and parsing the target model is conducive to the user's understanding of the target model. For example, when it is determined that the algorithm index and the business index are both greater than or equal to the corresponding threshold, the model structure of the target model is analyzed to generate a model structure diagram. In addition, when it is determined that the algorithm index and the business index are both greater than or equal to the corresponding threshold, the feature dependency relationship of the target model is analyzed to generate a feature dependency diagram.

[0084] Reference below Figure 7 , the model training method shown above is further described in conjunction with a specific embodiment. Those skilled in the art can understand that the following exemplary embodiments are only used to understand the present disclosure, and the present disclosure is not limited thereto.

[0085] Exemplarily, in this embodiment, the first object may include, for example, the entrusted party's salesperson (AM), solution architect (SA), product manager (PM), project engineer (RD), etc. The second object may include, for example, the entrusted party's business personnel.

[0086] Figure 7 A schematic diagram of a model training method according to another embodiment of the present disclosure is schematically shown.

[0087] exist Figure 7It is shown in the figure that the sales staff, solution architects, product managers and project engineers of the entrusted party can hold a voice conference or video conference with the business personnel of the entrusting party through the project management server of the deep learning model training project to meet the needs. The project management server can recognize the content of the voice conference or video conference and generate corresponding communication minutes, which are stored in the project management server. Product managers and project engineers can access the project management server to obtain the communication minutes and start the solution design based on the communication minutes.

[0088] When designing a solution, product managers and project engineers can evaluate the feasibility of the client's requirements. If the evaluation is not feasible, the requirements need to be sent back for re-communication. If the evaluation is feasible, a solution is produced and uploaded to the project management server to start preparations for entry.

[0089] When preparing for entry, the sales staff and the business staff of the entrusting party upload the signature information of the entrustee and the entrusting party respectively to sign relevant agreements, such as security agreements, confidentiality agreements, legal agreements, etc. Then the project management server can verify the signature information of the entrustee and the entrusting party, and after the verification is passed, apply to open the corresponding hardware and software permissions for the entrusting party, that is, configure the permission information.

[0090] Then, data preparation is started. Product managers and project engineers can send demand requests to the client's business personnel through the project management server. After receiving the demand request, the client's business personnel prepare the original sample data and upload the original sample data to the project management server for modeling.

[0091] The project management server verifies the authorization information of the entrusted party and the original sample data provided by the entrusting party. When it is determined that both the authorization information and the original sample data have passed the verification, the original sample data can be preprocessed to obtain the target sample data. Then, the deep learning model can be trained using the target sample data to obtain the target model. If at least one of the authorization information and the original sample data has not passed the verification, the process returns to the demand communication stage.

[0092] After the modeling data is ready, the model building is started. The project management server can process the original sample data and output a wide data table. Then, data exploration can be performed on the wide data table, and a sample analysis report can be output based on the wide data table. Next, feature engineering can be performed to output modelable features based on the wide data table, that is, feature files. Then, the feature files are input into the deep learning model to train the deep learning model and output the target model.

[0093] After the model file is produced, the target model can be evaluated to produce algorithm indicators and business indicators. The indicators can be verified by the project management server or manually. If all indicators meet the standards, the model interpretation stage can be entered. In the model interpretation stage, the project management server or project engineer can analyze the structure and feature dependencies of the target model to generate a model structure diagram and a feature dependency diagram. If at least one indicator does not meet the standards, the target model can be optimized, and then the data processing step can be returned to retrain the target model. Exemplarily, in this embodiment, the model optimization is performed up to 3 times, and if all 3 times do not meet the standards, the iteration is terminated.

[0094] After the model is evaluated, the target model can be published. Exemplarily, in this embodiment, two online methods can be adopted: offline prediction and online prediction. For the offline prediction method, the client can upload the sample to the project management server, and the project management server uses the target model to output the offline prediction result based on the sample for the client to download. For the online prediction method, an access interface (API) can be set and the interface document can be provided to the client. The client can access the corresponding access interface according to the interface document to use the target model.

[0095] According to an embodiment of the present disclosure, after the model is launched, the project management server may also regularly output evaluation reports of online model indicators and business indicators.

[0096] According to the embodiments of the present disclosure, the project management server can coordinate the work of the entrusting party and the entrusting party, prevent duplication of work, prevent process disruption, and improve the efficiency of training deep learning models.

[0097] Figure 8 A block diagram of a model training device according to an embodiment of the present disclosure is schematically shown.

[0098] like Figure 8 As shown, the model training device 800 includes a verification module 810, a data processing module 820 and a model training module 830.

[0099] The verification module 810 is used to verify the permission information of the first object and the original sample data of the second object.

[0100] The data processing module 820 is used to determine the target sample data according to the original sample data when it is determined that both the authority information and the original sample data have passed the verification.

[0101] The model training module 830 is used to train the deep learning model using the target sample data to obtain the target model.

[0102] According to an embodiment of the present disclosure, the above-mentioned device may also include a signature receiving module, a signature verification module and a permission configuration module. The signature receiving module is used to receive a first signature information from a first operating platform and a second signature information from a second operating platform, wherein the first operating platform corresponds to the first object and the second operating platform corresponds to the second object. The signature verification module is used to verify the first signature information and the second signature information. The permission configuration module is used to configure the permission information of the first object when both the first signature information and the second signature information pass the verification.

[0103] According to an embodiment of the present disclosure, the above-mentioned device may further include a demand receiving module, a request sending module and a data receiving module. Among them, the demand receiving module is used to generate a demand request in response to receiving data demand information from the first operating platform according to the data demand information. The request sending module is used to send the demand request to the second operating platform. The data receiving module is used to receive the original sample data from the second operating platform.

[0104] According to an embodiment of the present disclosure, the original sample data may include multiple data tables. The data processing module includes a field extraction submodule and a data wide table determination submodule. The field extraction submodule is used to extract multiple target fields from the multiple data tables. The data wide table determination submodule is used to determine the data wide table as the target sample data according to the multiple target fields.

[0105] According to an embodiment of the present disclosure, the model training module may include a feature engineering submodule, an input submodule and an adjustment module. The feature engineering submodule is used to perform data preprocessing on the target sample data to obtain a feature file. The input submodule is used to input the deep learning model using the feature file to obtain an output result. The adjustment submodule is used to adjust the parameters of the deep learning model according to the output result.

[0106] According to an embodiment of the present disclosure, the above-mentioned device may also include an interface generation module, an interface generation module, a model output module and an interface sending module. Among them, the interface generation module is used to generate an access interface corresponding to the target model. The model receiving module is used to receive target input data from the second operating platform through the access interface. The model output module is used to determine the target output data corresponding to the target input data using the target model. The interface sending module is used to send the target output data to the second operating platform through the access interface.

[0107] According to an embodiment of the present disclosure, the above-mentioned device may further include a data exploration module and a sample adjustment module. The data exploration module is used to perform data exploration on the target sample data to obtain sample distribution information. The sample adjustment module is used to adjust the target sample data according to the sample distribution information.

[0108] According to an embodiment of the present disclosure, the above-mentioned device may further include a model evaluation module and a model optimization module. The model evaluation module is used to perform model evaluation on the target model to obtain algorithm indicators and business indicators. The model optimization module is used to optimize the target model when it is determined that at least one of the algorithm indicators and the business indicators is less than the corresponding threshold.

[0109] According to an embodiment of the present disclosure, the above-mentioned device may also include a first model interpretation module, which is used to analyze the model structure of the target model and generate a model structure diagram when it is determined that the algorithm indicator and the business indicator are both greater than or equal to the corresponding threshold value.

[0110] According to an embodiment of the present disclosure, the above-mentioned device may also include a second model interpretation module, which is used to analyze the feature dependency relationship of the target model and generate a feature dependency graph when it is determined that the algorithm indicator and the business indicator are both greater than or equal to the corresponding threshold value.

[0111] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0112] Fig. 9 A block diagram of an example electronic device 900 that can be used to implement an embodiment of the present disclosure is schematically shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0113] like Fig. 9 As shown, the device 900 includes a computing unit 901, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 902 or a computer program loaded from a storage unit 908 into a random access memory (RAM) 903. In the RAM 903, various programs and data required for the operation of the device 900 can also be stored. The computing unit 901, the ROM 902, and the RAM 903 are connected to each other via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0114] A number of components in the device 900 are connected to the I / O interface 905, including: an input unit 906, such as a keyboard, a mouse, etc.; an output unit 907, such as various types of displays, speakers, etc.; a storage unit 908, such as a disk, an optical disk, etc.; and a communication unit 909, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 909 allows the device 900 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0115] The computing unit 901 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 901 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 901 performs the various methods and processes described above, such as the model training method. For example, in some embodiments, the model training method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 908. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 900 via the ROM 902 and / or the communication unit 909. When the computer program is loaded into the RAM 903 and executed by the computing unit 901, one or more steps of the model training method described above may be performed. Alternatively, in other embodiments, the computing unit 901 may be configured to perform the model training method in any other appropriate manner (e.g., by means of firmware).

[0116] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0117] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0118] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0119] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0120] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0121] A computer system may include clients and servers. Clients and servers are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship to each other.

[0122] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0123] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A model training method, comprising: Verify the permission information of the first object and the original sample data of the second object; wherein the first object is the entrusted party, and the second object is: the entrusting party who entrusts the first object to train the deep learning model; When it is determined that both the authority information and the original sample data have passed the verification, determining the target sample data according to the original sample data; and Using the target sample data to train a deep learning model to obtain a target model, The permission information of the first object is used to indicate whether the first object has the permission to use the hardware and software corresponding to the model training; the verification of the original sample data includes verifying whether the original sample data contains all the fields required for training; The method further comprises: Receiving first signature information from a first operating platform and second signature information from a second operating platform, wherein the first operating platform corresponds to the first object and the second operating platform corresponds to the second object; Verifying the first signature information and the second signature information; When both the first signature information and the second signature information pass verification, configuring permission information of the first object; Among them, the first signature information includes the signature image and / or seal image of the first object; the second signature information includes the signature image and / or seal image of the second object; the first signature information and the second signature information are both verified to indicate that the first object and the second object have reached a delegation agreement.

2. The method according to claim 1, further comprising: In response to receiving the data demand information from the first operating platform, generating a demand request according to the data demand information; Sending the demand request to the second operating platform; as well as Receive raw sample data from the second operating platform.

3. The method according to claim 1, wherein: The original sample data includes a plurality of data tables; and determining the target sample data according to the original sample data includes: Extracting multiple target fields from the multiple data tables; and According to the multiple target fields, a data wide table is determined as the target sample data.

4. The method according to claim 3, wherein: The method of using the target sample data to train a deep learning model to obtain a target model includes: Performing data preprocessing on the target sample data to obtain a feature file; Using the feature file to input the deep learning model to obtain an output result; and Adjust the parameters of the deep learning model according to the output results.

5. The method according to any one of claims 1 to 4, further comprising: Generate an access interface corresponding to the target model; Receiving target input data from the second operating platform via the access interface; Determining target output data corresponding to the target input data using the target model; as well as The target output data is sent to the second operating platform through the access interface.

6. The method according to any one of claims 1 to 4, further comprising: Performing data exploration on the target sample data to obtain sample distribution information; as well as The target sample data is adjusted according to the sample distribution information.

7. The method according to any one of claims 1 to 4, further comprising: Performing model evaluation on the target model to obtain algorithm indicators and business indicators; as well as When it is determined that at least one of the algorithm indicator and the business indicator is less than a corresponding threshold, the target model is optimized.

8. The method according to claim 7, further comprising: When it is determined that the algorithm indicator and the business indicator are both greater than or equal to the corresponding thresholds, the model structure of the target model is analyzed to generate a model structure diagram.

9. The method according to claim 7, further comprising: When it is determined that the algorithm indicator and the business indicator are both greater than or equal to the corresponding thresholds, the feature dependency relationship of the target model is analyzed to generate a feature dependency graph.

10. A model training device, comprising: A verification module, used to verify the permission information of the first object and the original sample data of the second object; wherein the first object is the entrusted party, and the second object is: the entrusting party who entrusts the first object to train the deep learning model on its behalf; a data processing module, configured to determine target sample data according to the original sample data when it is determined that both the permission information and the original sample data have passed verification; and A model training module is used to train a deep learning model using the target sample data to obtain a target model. The permission information of the first object is used to indicate whether the first object has the permission to use the hardware and software corresponding to the model training; the verification of the original sample data includes verifying whether the original sample data contains all the fields required for training; the device also includes: A signature receiving module, configured to receive first signature information from a first operating platform and second signature information from a second operating platform, wherein the first operating platform corresponds to the first object and the second operating platform corresponds to the second object; A signature verification module, used to verify the first signature information and the second signature information; A permission configuration module, configured to configure permission information of the first object when both the first signature information and the second signature information pass verification; Among them, the first signature information includes the signature image and / or seal image of the first object; the second signature information includes the signature image and / or seal image of the second object; the first signature information and the second signature information are both verified to indicate that the first object and the second object have reached a delegation agreement.

11. The apparatus according to claim 10, further comprising: a demand receiving module, configured to generate a demand request in response to receiving data demand information from the first operating platform and according to the data demand information; A request sending module, used for sending the demand request to the second operating platform; as well as The data receiving module is used to receive the original sample data from the second operating platform.

12. The device according to claim 10, wherein: The original sample data includes multiple data tables; The data processing module comprises: A field extraction submodule, used to extract multiple target fields from the multiple data tables; as well as The data wide table determination submodule is used to determine a data wide table as the target sample data according to the multiple target fields.

13. The device according to claim 12, wherein: The model training module includes: A feature engineering submodule is used to perform data preprocessing on the target sample data to obtain a feature file; An input submodule, used to input the deep learning model using the feature file to obtain an output result; and An adjustment submodule is used to adjust the parameters of the deep learning model according to the output result.

14. The apparatus according to any one of claims 10 to 13, further comprising: An interface generation module, used to generate an access interface corresponding to the target model; A model receiving module, used for receiving target input data from the second operating platform through the access interface; A model output module, used to determine target output data corresponding to the target input data using the target model; as well as An interface sending module is used to send the target output data to the second operating platform through the access interface.

15. The apparatus according to any one of claims 10 to 13, further comprising: A data exploration module, used to perform data exploration on the target sample data to obtain sample distribution information; as well as The sample adjustment module is used to adjust the target sample data according to the sample distribution information.

16. The apparatus according to any one of claims 10 to 13, further comprising: A model evaluation module is used to perform model evaluation on the target model to obtain algorithm indicators and business indicators; as well as The model optimization module is used to optimize the target model when it is determined that at least one of the algorithm indicator and the business indicator is less than a corresponding threshold.

17. The apparatus according to claim 16, further comprising: The first model interpretation module is used to analyze the model structure of the target model and generate a model structure diagram when it is determined that the algorithm indicator and the business indicator are both greater than or equal to the corresponding threshold value.

18. The apparatus according to claim 16, further comprising: The second model interpretation module is used to analyze the feature dependency relationship of the target model and generate a feature dependency graph when it is determined that the algorithm indicator and the business indicator are both greater than or equal to the corresponding threshold value.

19. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 9.

20. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-9.

21. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 9 are implemented.

Images