Model training method and device, prediction probability information generation method and device and electronic equipment

Abstract

The embodiment of the invention discloses a model training method, a prediction probability information generation method, a prediction probability information generation device and electronic equipment. A specific embodiment of the method comprises the steps of obtaining a training data set; inputting the variable data set into an initial feature data vector representation model to output an initial variable vector set; performing model training on each initial discrimination model in an initial discrimination model set included in the initial target value payment mode probability generation model to obtain a discrimination model set; and according to the initial variable vector set, the actual value reduction information set and the actual value payment mode information set, performing model training on an initial generative model set and an initial feature data vector representation model included in the trained target value payment mode probability generation model to obtain the target value payment mode probability generation model. The implementation mode is related to artificial intelligence, and the target value payment mode prediction probability information can be accurately generated by using the trained target value payment mode probability generation model.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of computer technology, and in particular to a model training method, a prediction probability information generation method, a device, and an electronic device. Background Art

[0002] At present, in the value circulation scenario of items, certain intervention means (for example, value push) can greatly influence the decision-making behavior of users. For the generation of the probability of the target value payment method, the method usually adopted is: to determine the predicted probability information of the target value payment method through the causal inference method (DeR-CFR, Decomposed Representations for Conterfactual Regression).

[0003] However, the inventors have found that when the above method is adopted, the following technical problems often occur:

[0004] The problem of inaccurate representation of individual causal effects leads to inaccurate prediction probability information of the generated target value payment method.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the inventive concept and therefore it may contain information that does not form the prior art that is already known in this country to a person of ordinary skill in the art. Summary of the invention

[0006] The content of this disclosure is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this disclosure is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.

[0007] Some embodiments of the present disclosure propose a model training method, a prediction probability information generation method, a device and an electronic device to solve the technical problems mentioned in the above background technology section.

[0008] In a first aspect, some embodiments of the present disclosure provide a model training method, comprising: obtaining a training data set, wherein the training data in the training data set includes: variable data, actual value reduction information, and actual value payment method information; inputting the variable data set into an initial feature data vector representation model included in an initial target value payment method probability generation model to output an initial variable vector set; based on the initial variable vector set, the actual value reduction information set, and the actual value payment method information set, performing model training on each initial discriminant model set included in the initial target value payment method probability generation model to obtain a discriminant model set, wherein the discriminant model set includes: a first value reduction information generation model, a first value payment method probability generation model, Generate a model and a second value payment method probability generation model; fix the model parameters of the above-mentioned first value reduction information generation model, the above-mentioned first value payment method probability generation model and the above-mentioned second value payment method probability generation model, and according to the above-mentioned initial variable vector set, the above-mentioned actual value reduction information set and the above-mentioned actual value payment method information set, perform model training on the initial generative model set and the above-mentioned initial feature data vector representation model included in the trained target value payment method probability generation model to obtain the target value payment method probability generation model, wherein the generative model set in the above-mentioned target value payment method probability generation model includes: a third value payment method probability generation model, a fourth value payment method probability generation model, and a second value reduction information generation model.

[0009] Optionally, the above-mentioned variable data includes: instrumental variable data, adjustment variable data, and confounding variable data; and the above-mentioned input of the variable data set into the initial target value payment method probability generation model included in the initial feature data vector representation model to output an initial variable vector set, including: inputting the instrumental variable data set, the adjustment variable data set and the confounding variable data set into the above-mentioned initial feature data vector representation model to generate an initial instrumental variable vector set, an initial adjustment variable vector set and an initial confounding variable vector set; vector set fusion of the above-mentioned initial instrumental variable vector set, the above-mentioned initial adjustment variable vector set and the above-mentioned initial confounding variable vector set to obtain an initial variable vector set.

[0010] Optionally, the above-mentioned initial discriminant model set included in the initial target value payment method probability generation model is trained according to the above-mentioned initial variable vector set, the actual value reduction information set and the actual value payment method information set to obtain the discriminant model set, including: selecting an initial adjustment variable vector from the above-mentioned initial adjustment variable vector set as the first initial adjustment variable vector, and performing the following first training step: inputting the above-mentioned first initial adjustment variable vector into the first initial value reduction information generation model included in the above-mentioned initial target value payment method probability generation model to generate the first initial value reduction information; determining the actual value reduction information corresponding to the above-mentioned first initial adjustment variable vector as the first actual value reduction information. information; determining first loss information for the first initial value reduction information and the first actual value reduction information; in response to determining that the first loss information is greater than the first numerical value, determining the first initial value reduction information generation model as the first value reduction information generation model; in response to determining that the first loss information is less than or equal to the first numerical value, updating the model parameters of the first initial value reduction information generation model according to the first loss information to obtain an updated first value reduction information generation model, and reselecting the first initial adjustment variable vector from the initial adjustment variable vector set, determining the updated first value reduction information generation model as the first initial value reduction information generation model, so as to continue to execute the first training step.

[0011] Optionally, the above-mentioned initial discriminant models in the initial discriminant model set included in the initial target value payment method probability generation model are trained according to the above-mentioned initial variable vector set, the actual value reduction information set and the actual value payment method information set to obtain the discriminant model set, including: selecting an initial confounding variable vector from the above-mentioned initial confounding variable vector set as the first initial confounding variable vector, and performing the following second training step: determining the actual value reduction information corresponding to the above-mentioned first initial confounding variable vector as the second actual value reduction information, and determining the initial instrumental variable vector corresponding to the above-mentioned first initial confounding variable vector as the first initial instrumental variable vector; generating a first actual value reduction vector for the above-mentioned second actual value reduction information; inputting the above-mentioned first initial confounding variable vector and the above-mentioned first actual value reduction vector into the first initial value payment method probability generation model to generate the first initial value payment method probability information; inputting the above-mentioned first initial instrumental variable vector, the above-mentioned first initial confounding variable vector and the above-mentioned first actual value reduction vector into the second initial value payment method probability generation model to generate the second initial value payment method probability information. method probability information; generate second loss information for the above-mentioned first initial value payment method probability information and the above-mentioned second initial value payment method probability information; in response to determining that the above-mentioned second loss information is greater than the second value, determine the above-mentioned first initial value payment method probability generation model as the first value payment method probability generation model, and determine the above-mentioned second initial value payment method probability generation model as the second value payment method probability generation model; in response to determining that the above-mentioned second loss information is less than or equal to the above-mentioned second value, according to the above-mentioned second loss information, update the model parameters of the above-mentioned first initial value payment method probability generation model and the second initial value payment method probability generation model to obtain the updated first value payment method probability generation model and the updated first value payment method probability generation model, and re-select the first initial confusion variable vector from the above-mentioned initial confusion variable vector set, use the above-mentioned updated first value payment method probability generation model as the first initial value payment method probability generation model, and use the above-mentioned updated second value payment method probability generation model as the second initial value payment method probability generation model, so as to continue to execute the above-mentioned second training step.

[0012] Optionally, the above-mentioned trained target value payment method probability generation model includes an initial generative model set and the above-mentioned initial feature data vector representation model, and a model training is performed to obtain a target value payment method probability generation model, including: selecting an initial confounding variable vector from the above-mentioned initial confounding variable vector set as a second initial confounding variable vector, and performing the following third training step: determining the initial adjustment variable vector, the initial instrumental variable vector, the corresponding actual value reduction vector and the corresponding actual value reduction information corresponding to the above-mentioned second initial confounding variable vector, as the second initial adjustment variable vector, the second initial instrumental variable vector, the third actual value reduction vector and the third actual value reduction information respectively; inputting the above-mentioned second initial confounding variable vector, the third actual value reduction vector and the above-mentioned second initial adjustment variable vector into the third initial value payment method probability generation model to generate the third initial value payment method probability information; inputting the above-mentioned second initial adjustment variable vector into the fourth initial value payment method probability generation model to generate the fourth initial value payment method probability information. method probability information; inputting the above-mentioned second initial instrumental variable vector into the second initial value reduction information generation model to generate the second initial value reduction information; generating third loss information for the above-mentioned third initial value payment method probability information and the corresponding actual value payment method information; generating fourth loss information for the above-mentioned fourth initial value payment method probability information and the corresponding actual value payment method information; generating fifth loss information corresponding to the above-mentioned second initial value reduction information and the above-mentioned third actual value reduction information; in response to determining that the above-mentioned third loss information is less than the third numerical value, the above-mentioned fourth loss information is less than the fourth numerical value, and the above-mentioned fifth loss information is less than the fifth numerical value, the above-mentioned third initial value payment method probability generation model, the above-mentioned fourth initial value payment method probability generation model, the above-mentioned second initial value reduction information generation model and the above-mentioned initial characteristic data vector representation model are respectively determined as the third value payment method probability generation model, the fourth value payment method probability generation model, the second value reduction information generation model and the characteristic data vector representation model.

[0013] Optionally, the above-mentioned target value payment method probability generation model is verified by the following steps: obtaining a user information set; for each value reduction value among the multiple value reduction values ​​included in the value reduction information, executing the following information generation steps: determining the value issuance value corresponding to each user information in the above-mentioned user information set; screening out a user information subset whose corresponding value issuance value is the target value from the above-mentioned user information set as the first user information subset; screening out a user information subset whose corresponding value issuance value is the above-mentioned value reduction value from the above-mentioned user information set as the second user information subset; fusing the above-mentioned first user information subset and the above-mentioned second user information subset to obtain a fused user information set; for each fused user information in the above-mentioned fused user information set, using the above-mentioned target value payment method probability generation model, determining the prediction probability of the first target value payment method of the above-mentioned fused user information under the above-mentioned value reduction value rate information and the predicted probability information of the second target value payment method of the above-mentioned fused user information under the above-mentioned target value; according to the information difference between the predicted probability information of the first target value payment method and the predicted probability information of the second target value payment method corresponding to each fused user information, the user information of each fused user information in the above-mentioned fused user information set is grouped to obtain at least one user information group; for each user information group in the above-mentioned at least one user information group, according to the information difference set corresponding to the above-mentioned user information group, the predicted average gain value corresponding to the above-mentioned user information group is determined; the actual average gain value for each user information group is determined to obtain at least one actual average gain value; the gain value of the at least one predicted average gain value obtained is compared with the at least one actual average gain value mentioned above to obtain comparison information; according to the obtained multiple comparison information, a model verification result for the above-mentioned target value payment method probability generation model is generated.

[0014] In a second aspect, some embodiments of the present disclosure provide a model training device, including: a first acquisition unit, configured to acquire a training data set, wherein the training data in the above-mentioned training data set includes: variable data, actual value reduction information and actual value payment method information; a first generation unit, configured to input the variable data set into an initial feature data vector representation model included in an initial target value payment method probability generation model to output an initial variable vector set; a first training unit, configured to perform model training on each initial discriminant model set included in the above-mentioned initial target value payment method probability generation model according to the above-mentioned initial variable vector set, actual value reduction information set and actual value payment method information set, to obtain a discriminant model set, wherein the above-mentioned discriminant model set includes: a first value reduction information generation model , a first value payment method probability generation model and a second value payment method probability generation model; a second training unit is configured to fix the model parameters of the above-mentioned first value reduction information generation model, the above-mentioned first value payment method probability generation model and the above-mentioned second value payment method probability generation model, and according to the above-mentioned initial variable vector set, the above-mentioned actual value reduction information set and the above-mentioned actual value payment method information set, perform model training on the initial generative model set and the above-mentioned initial feature data vector representation model included in the trained target value payment method probability generation model to obtain the target value payment method probability generation model, wherein the generative model set in the above-mentioned target value payment method probability generation model includes: a third value payment method probability generation model, a fourth value payment method probability generation model, and a second value reduction information generation model.

[0015] Optionally, the variable data include: instrumental variable data, adjustment variable data, and confounding variable data; and the first generating unit may be configured to: select an initial adjustment variable vector from the initial adjustment variable vector set as the first initial adjustment variable vector, and perform the following first training step: input the first initial adjustment variable vector into the first initial value reduction information generating model included in the initial target value payment method probability generating model to generate the first initial value reduction information; determine the actual value reduction information corresponding to the first initial adjustment variable vector as the first actual value reduction information; determine the actual value reduction information corresponding to the first initial value reduction information and the first actual value reduction information; In response to determining that the first loss information is greater than the first numerical value, determining the first initial value reduction information generation model as the first value reduction information generation model; in response to determining that the first loss information is less than or equal to the first numerical value, updating the model parameters of the first initial value reduction information generation model according to the first loss information to obtain the updated first value reduction information generation model, and reselecting the first initial adjustment variable vector from the initial adjustment variable vector set, determining the updated first value reduction information generation model as the first initial value reduction information generation model, so as to continue to execute the first training step.

[0016] Optionally, the first training unit can be configured to: select an initial confounding variable vector from the above-mentioned initial confounding variable vector set as the first initial confounding variable vector, and perform the following second training step: determine the actual value reduction information corresponding to the above-mentioned first initial confounding variable vector as the second actual value reduction information, and determine the initial instrumental variable vector corresponding to the above-mentioned first initial confounding variable vector as the first initial instrumental variable vector; generate a first actual value reduction vector for the above-mentioned second actual value reduction information; input the above-mentioned first initial confounding variable vector and the above-mentioned first actual value reduction vector into the first initial value payment method probability generation model to generate the first initial value payment method probability information; input the above-mentioned first initial instrumental variable vector, the above-mentioned first initial confounding variable vector and the above-mentioned first actual value reduction vector into the second initial value payment method probability generation model to generate the second initial value payment method probability information; generate the above-mentioned first initial instrumental variable vector, the above-mentioned first initial confounding variable vector and the above-mentioned first actual value reduction vector into the second initial value payment method probability generation model to generate the second initial value payment method probability information; generate the above-mentioned first initial value payment method probability information and the above-mentioned second initial value payment method probability information. in response to determining that the second loss information is greater than the second value, determining the first initial value payment method probability generation model as the first value payment method probability generation model, and determining the second initial value payment method probability generation model as the second value payment method probability generation model; in response to determining that the second loss information is less than or equal to the second value, updating the model parameters of the first initial value payment method probability generation model and the second initial value payment method probability generation model according to the second loss information to obtain an updated first value payment method probability generation model and an updated first value payment method probability generation model, and reselecting the first initial confusion variable vector from the initial confusion variable vector set, using the updated first value payment method probability generation model as the first initial value payment method probability generation model, and using the updated second value payment method probability generation model as the second initial value payment method probability generation model, so as to continue to execute the second training step.

[0017] Optionally, the second training unit can be configured to: select an initial confounding variable vector from the above-mentioned initial confounding variable vector set as the second initial confounding variable vector, and perform the following third training step: determine the initial adjustment variable vector, the initial instrumental variable vector and the corresponding actual value reduction information corresponding to the above-mentioned second initial confounding variable vector, as the second initial adjustment variable vector, the second initial instrumental variable vector and the third actual value reduction information respectively; input the above-mentioned second initial confounding variable vector and the above-mentioned second initial adjustment variable vector into the third initial value payment method probability generation model to generate the third initial value payment method probability information; input the above-mentioned second initial adjustment variable vector into the fourth initial value payment method probability generation model to generate the fourth initial value payment method probability information; input the above-mentioned second initial instrumental variable vector into the second initial value reduction information generation model to generate the second initial value reduction information; generate third loss information for the third initial value payment method probability information and the corresponding actual value payment method information; generate fourth loss information for the fourth initial value payment method probability information and the corresponding actual value payment method information; generate fifth loss information corresponding to the second initial value reduction information and the third actual value reduction information; in response to determining that the third loss information is less than the third numerical value, the fourth loss information is less than the fourth numerical value, and the fifth loss information is less than the fifth numerical value, the third initial value payment method probability generation model, the fourth initial value payment method probability generation model, the second initial value reduction information generation model, and the initial feature data vector representation model are respectively determined as the third value payment method probability generation model, the fourth value payment method probability generation model, the second value reduction information generation model, and the feature data vector representation model.

[0018] In a third aspect, some embodiments of the present disclosure provide a method for generating predicted probability information, including: obtaining variable data for a target user and multiple value reduction values ​​included in value reduction information; for each of the multiple value reduction values, inputting the variable data and the value reduction value into a pre-trained target value payment method probability generation model to generate target value payment method predicted probability information for the value reduction value, wherein the target value payment method probability generation model is generated based on the model training method of the first aspect of the present disclosure; selecting a value reduction value whose corresponding target value payment method predicted probability information meets a preset information condition from the multiple value reduction values ​​as the target value reduction value; and pushing the reduction information corresponding to the target value reduction value to the user terminal corresponding to the target user.

[0019] Optionally, for each of the above-mentioned multiple value reduction values, the above-mentioned variable data and the above-mentioned value reduction value are input into a pre-trained target value payment method probability generation model to generate target value payment method prediction probability information for the above-mentioned value reduction value, including: for each of the above-mentioned multiple value reduction values, the above-mentioned variable data and the above-mentioned value reduction value are input into a third value payment method probability generation model included in the target value payment method probability generation model to generate target value payment method prediction probability information for the above-mentioned value reduction value.

[0020] In a fourth aspect, some embodiments of the present disclosure provide a prediction probability information generating device, comprising: a second acquisition unit, configured to acquire variable data for a target user and a plurality of value reduction values ​​included in value reduction information; a second generation unit, configured to input the variable data and the value reduction value into a pre-trained target value payment method probability generation model for each of the plurality of value reduction values, so as to generate target value payment method prediction probability information for the value reduction value, wherein the target value payment method probability generation model is generated based on the model training method of the first aspect of the present disclosure; a screening unit, configured to screen out, from the plurality of value reduction values, a value reduction value whose corresponding target value payment method prediction probability information meets a preset information condition, as a target value reduction value; a push unit, configured to push the reduction information corresponding to the target value reduction value to a user terminal corresponding to the target user.

[0021] Optionally, the second generating unit may be configured to: for each of the above-mentioned multiple value reduction values, input the above-mentioned variable data and the above-mentioned value reduction value into a third value payment method probability generation model included in the target value payment method probability generation model, so as to generate target value payment method predicted probability information for the above-mentioned value reduction value.

[0022] In a fifth aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any one of the implementation methods in the first and third aspects.

[0023] In a sixth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method described in any one of the implementation modes of the first aspect and the third aspect is implemented.

[0024] In a seventh aspect, some embodiments of the present disclosure provide a computer program product, including a computer program, which, when executed by a processor, implements the method described in any one of the implementation modes of the first and third aspects above.

[0025] The above-mentioned embodiments of the present disclosure have the following beneficial effects: by using the trained target value payment method probability generation model through the model training method of some embodiments of the present disclosure, the target value payment method prediction probability information can be accurately generated. Specifically, the reason why the relevant target value payment method prediction probability information is not accurate enough is that the problem of inaccurate representation of individual causal effects leads to inaccurate generated target value payment method prediction probability information. Based on this, the model training method of some embodiments of the present disclosure first obtains a training data set, wherein the training data in the above training data set includes: variable data, actual value reduction information and actual value payment method information. Here, the obtained variable data, real-time value reduction information and real-time value payment method information are used as input data and data labels for model training in the future. Then, the variable data set is input into the initial feature data vector representation model included in the initial target value payment method probability generation model to output the initial variable vector set. Here, the variable data in the variable data set can be converted into a vector form through the initial feature data vector representation model, so as to facilitate the subsequent input into the initial target value payment method probability generation model. Next, according to the above-mentioned initial variable vector set, actual value reduction information set and actual value payment method information set, each initial discriminant model in the initial discriminant model set included in the above-mentioned initial target value payment method probability generation model is trained to obtain a discriminant model set, wherein the above-mentioned discriminant model set includes: a first value reduction information generation model, a first value payment method probability generation model and a second value payment method probability generation model. Here, by first training each discriminant model in the initial discriminant model set, an accurate discriminant model set is trained to ensure that it will serve as an auxiliary training model for the initial generative model set in the future, and help the model training of the initial generative model set. In addition, the first value reduction information generation model, the first value payment method probability generation model and the second value payment method probability generation model can be discriminant models set based on variable causal inference, which can accurately characterize individual causal effects. Therefore, by setting the first value reduction information generation model, the first value payment method probability generation model and the second value payment method probability generation model as discriminators to assist the training of the subsequent third value payment method probability generation model, the fourth value payment method probability generation model and the second value reduction information generation model, the accuracy of model training can be greatly improved.Finally, the model parameters of the first value reduction information generation model, the first value payment method probability generation model and the second value payment method probability generation model are fixed, and the initial generative model set and the initial feature data vector representation model included in the trained target value payment method probability generation model are trained according to the initial variable vector set, the actual value reduction information set and the actual value payment method information set to obtain the target value payment method probability generation model, wherein the generative model set in the target value payment method probability generation model includes: the third value payment method probability generation model, the fourth value payment method probability generation model and the second value reduction information generation model. Here, the third value payment method probability generation model, the fourth value payment method probability generation model and the second value reduction information generation model are also generative models set based on causal inference, which can accurately generate the corresponding individual causal effect information. With the assistance of model training of the discriminant model set, the target value payment method probability generation model can be accurately obtained. In summary, by first training the discriminative model set and then training the generative model set, the individual causal effects can be fully characterized, so that the trained target value payment method probability generation model can be used to accurately generate the target value payment method prediction probability information. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that components and elements are not necessarily drawn to scale.

[0027] Figure 1-Figure 2 is a schematic diagram of an application scenario of a model training method according to some embodiments of the present disclosure;

[0028] Figure 3 is a flowchart of some embodiments of the model training method according to the present disclosure;

[0029] Figure 4 is a flowchart of some embodiments of the method for generating prediction probability information according to the present disclosure;

[0030] Figure 5 is a flowchart of other embodiments of the method for generating prediction probability information according to the present disclosure;

[0031] Figure 6 is a schematic diagram of the structure of some embodiments of the model training device according to the present disclosure;

[0032] Figure 7 is a schematic diagram of the structure of some embodiments of the prediction probability information generating device according to the present disclosure;

[0033] Figure 8 It is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0034] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0035] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure can be combined with each other.

[0036] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0037] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0038] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0039] With regard to the collection, storage, and use of user information (such as actual value reduction information and actual value payment method information) involved in this disclosure, before performing the corresponding operations, the relevant organizations or individuals shall fulfill their obligations, including conducting information security impact assessments, fulfilling the obligation to inform the information subject, and obtaining the authorization and consent of the information subject in advance.

[0040] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0041] Figure 1-Figure 2 It is a schematic diagram of an application scenario of the model training method according to some embodiments of the present disclosure.

[0042] exist Figure 1-Figure 2In the application scenario, first, the electronic device 101 can obtain a training data set 102. The training data in the training data set 102 includes: variable data, actual value reduction information, and actual value payment method information. Then, the electronic device 101 can input the variable data set 103 into the initial feature data vector representation model 107 included in the initial target value payment method probability generation model 106 to output an initial variable vector set 108. Next, the electronic device 101 can perform model training on each initial discriminant model set 109 included in the initial target value payment method probability generation model 106 according to the initial variable vector set 108, the actual value reduction information set 104, and the actual value payment method information set 105 to obtain a discriminant model set 112. The discriminant model set 112 includes: a first value reduction information generation model 1121, a first value payment method probability generation model 1122, and a second value payment method probability generation model 1123. In this application scenario, the initial target value payment method probability generation model 106 further includes: an initial generative model set 110. The initial generative model set 110 may include: a third initial value payment method probability generation model 1101, a fourth initial value payment method probability generation model 1102, and a second initial value reduction information generation model 1103. Finally, the electronic device 101 may fix the model parameters of the first value reduction information generation model 1121, the first value payment method probability generation model 1122, and the second value payment method probability generation model 1123, and perform model training on the initial generative model set 110 and the initial feature data vector representation model 107 included in the trained target value payment method probability generation model 111 according to the initial variable vector set 108, the actual value reduction information set 104, and the actual value payment method information set 105, to obtain a target value payment method probability generation model 113. The generative model set 115 in the target value payment method probability generation model 113 includes: a third value payment method probability generation model 1151, a fourth value payment method probability generation model 1152, and a second value reduction information generation model 1153. In this application scenario, the target value payment method probability generation model 113 also includes: a feature data vector representation model 114.

[0043] It should be noted that the electronic device 101 can be hardware or software. When the electronic device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or it can be implemented as a single server or a single terminal device. When the electronic device is embodied as software, it can be installed in the hardware devices listed above. It can be implemented as multiple software or software modules for providing distributed services, for example, or it can be implemented as a single software or software module. No specific limitation is made here.

[0044] It should be understood that Figure 1-Figure 2 The number of electronic devices in the embodiment is only for illustration. Any number of electronic devices may be provided according to implementation requirements.

[0045] Continue to refer Figure 3 , shows a process 300 of some embodiments of the model training method according to the present disclosure. The model training method comprises the following steps:

[0046] Step 301, obtaining a training data set.

[0047] In some embodiments, the execution entity of the above model training method (for example Figure 1 The electronic device 101 shown can obtain a training data set through a wired connection or a wireless connection. Among them, the training data in the above training data set includes: variable data, actual value reduction information and actual value payment method information. Among them, the training data in the training data set can be data used for subsequent model training. The variable data can be feature data related to the user. Specifically, the feature data can be feature information corresponding to the user feature variable. For example, for the e-commerce field, the variable data can include but is not limited to at least one of the following: user historical payment preference information, user shopping information, and user identification information. For example, the user historical payment preference information can be the user's preference information for using the target payment method to pay value. For example, the target payment method can be the target advance payment method. The actual value reduction information can be the value reduction transformation information of the corresponding item. For the e-commerce scenario, the actual value reduction information can be the actual coupon used, or it can be a payment instant discount coupon. The actual value payment method information can be the payment method information for the actual value payment. Specifically, the payment method information can be the identification information of the payment method. The payment method can be various prepayment methods. There is a data correspondence between the variable data, actual value reduction information and actual value payment method information included in each training data.

[0048] Step 302 , inputting the variable data set into the initial feature data vector representation model included in the initial target value payment method probability generation model to output an initial variable vector set.

[0049] In some embodiments, the execution subject may input the variable data set into the initial feature data vector representation model included in the initial target value payment method probability generation model to output the initial variable vector set. Among them, the initial target value payment method probability generation model may be a target value payment method probability generation model for which the model has not yet been trained. The target value payment method probability generation model may be a model for generating target value payment method prediction probability information. The target value payment method prediction probability information may be probability information under the premise that the value payment method used by the user is the target value payment method. The value payment method may be a payment method for value payment. The target value payment method may be a predetermined value payment method. For example, the target value payment method may be a payment method for value payment using a target application (Application). Among them, the target value payment method prediction probability information is a numerical value between 0 and 1. The larger the target value payment method prediction probability information, the more likely the corresponding user is to use the target value payment method as a payment method. The initial feature data vector representation model may be a feature data vector representation model for which the model has not yet been trained. The feature data vector representation model may be a model that converts feature data into corresponding vectors to represent the semantic content of feature data. Specifically, the feature data may be user feature data. There is a one-to-one correspondence between the initial variable vector in the initial variable vector set and the variable data in the variable data set. The variable vector can represent the data semantic content of the corresponding variable data. In practice, the probability generation model of the initial target value payment method can be a multi-layer serially connected temporal neural network model. The initial feature data vector representation model can be a Bert encoding model.

[0050] In some optional implementations of some embodiments, the variable data include: instrumental variable data, adjustment variable data, and confounding variable data. Among them, instrumental variable data may be variable content corresponding to instrumental variables (Instrumental Variables). Adjustment variable data may be variable content corresponding to adjustment variables (Adjustment Variables). Confounding variable data may be variable content corresponding to confounding variables (Confounders). Instrumental variables may be characteristic variables associated with intervention information (Treatment) and related to user characteristics. Instrumental variables may be determinant variable factors of intervention information. Intervention information may be information on intervention means. Specifically, intervention information may be information on coupons issued or information on instant discount coupons for payment. Adjustment variable data may be characteristic variables associated with target value payment methods and related to user characteristics. Adjustment variables may be determinant variable factors of target value payment methods. Confounding variable data may be characteristic variables associated with target value payment methods and intervention information and related to user characteristics. Confounding variables may be determinant variable factors of target value payment methods and intervention information.

[0051] Optionally, the step of inputting the variable data set into the initial feature data vector representation model included in the initial target value payment method probability generation model to output the initial variable vector set may include the following steps:

[0052] In the first step, the execution subject may input the instrumental variable data set, the adjustment variable data set and the confounding variable data set into the initial feature data vector representation model to generate an initial instrumental variable vector set, an initial adjustment variable vector set and an initial confounding variable vector set. The initial instrumental variable vector may represent the data semantic content of the instrumental variable data. The initial confounding variable vector may represent the data semantic content of the confounding variable data. The initial adjustment variable vector may represent the data semantic content of the adjustment variable data.

[0053] In a second step, the execution entity may perform vector set fusion on the initial instrumental variable vector set, the initial adjustment variable vector set and the initial confounding variable vector set to obtain an initial variable vector set.

[0054] As an example, the execution entity may fuse the initial instrumental variable vector set, the initial adjustment variable vector set, and the initial confounding variable vector set in a one-to-one correspondence to generate a fused vector as the initial variable vector to obtain an initial variable vector set.

[0055] Optionally, the initial feature data vector representation model includes: an instrumental variable data vector representation model for instrumental variable data, an adjusted variable data vector representation model for adjusted variable data, and a confused variable data vector representation model for confused variable data. In practice, the instrumental variable data vector representation model can be a Bert pre-trained model. The adjusted variable data vector representation model can be a word embedding model. The confused variable data vector representation model can be a Transformer encoding model.

[0056] And the execution subject may input the instrumental variable data set, the adjustment variable data set and the confounding variable data set into the initial feature data vector characterization model to generate an initial instrumental variable vector set, an initial adjustment variable vector set and an initial confounding variable vector set, including the following steps:

[0057] The first step is to input each instrumental variable data in the above instrumental variable data set into the instrumental variable data vector representation model to generate an initial instrumental variable vector and obtain an initial instrumental variable vector set.

[0058] In the second step, each adjustment variable data in the above adjustment variable data set is input into the adjustment variable data vector representation model to generate an initial adjustment variable vector, thereby obtaining an initial adjustment variable vector set.

[0059] The third step is to input each confounding variable data in the above confounding variable data set into the confounding variable data vector representation model to generate an initial confounding variable vector and obtain an initial confounding variable vector set.

[0060] Step 303, based on the above-mentioned initial variable vector set, actual value reduction information set and actual value payment method information set, each initial discriminant model in the initial discriminant model set included in the above-mentioned initial target value payment method probability generation model is trained to obtain a discriminant model set.

[0061] In some embodiments, the execution subject may perform model training on each initial discriminant model in the initial discriminant model set included in the initial target value payment method probability generation model according to the initial variable vector set, the actual value reduction information set and the actual value payment method information set to obtain a discriminant model set. The initial discriminant model in the initial discriminant model set may be a discriminant model whose model training has not yet been completed. Among them, the discriminant model set includes: a first value reduction information generation model, a first value payment method probability generation model and a second value payment method probability generation model. The first value reduction information generation model may be a model for generating value reduction information based on the initial variable vector set. In practice, the first value reduction information generation model may be a recurrent neural network model. The first value payment method probability generation model may be a model for generating probability information of using the target value payment method based on the first initial variable vector and the corresponding actual value reduction information. In practice, the first value payment method probability generation model may be a recurrent neural network model with multiple layers of serial connections. The second value payment method probability generation model may be a model for generating probability information of using the target value payment method based on the second initial variable vector and the corresponding actual value reduction information. In practice, the second value payment method probability generation model can be a multi-layer serially connected recurrent neural network model. The vector content of the first initial variable vector corresponding to the first value payment method probability generation model is different from the vector content of the first initial variable vector corresponding to the second value payment method probability generation model. Correspondingly, the network structure of the first value payment method probability generation model can also be different from the network structure of the second value payment method probability generation model. Specifically, the number of network layers corresponding to the two is different.

[0062] In some optional implementations of some embodiments, the above-mentioned initial discriminant model set included in the initial discriminant model set of the initial target value payment method probability generation model is trained according to the above-mentioned initial variable vector set, the actual value reduction information set and the actual value payment method information set to obtain the discriminant model set, which may include the following steps:

[0063] In the first step, an initial adjustment variable vector is selected from the above initial adjustment variable vector set as the first initial adjustment variable vector, and the following first training step is performed:

[0064] In sub-step 1, the execution entity may input the first initial adjustment variable vector into the first initial value reduction information generation model included in the initial target value payment method probability generation model to generate the first initial value reduction information. The first initial value reduction information generation model may be a first value reduction information generation model whose training has not yet been completed. The first value reduction information generation model may be a model for predicting value reduction information based on adjustment variable data.

[0065] In sub-step 2, the execution entity may determine actual value reduction information corresponding to the initial adjustment variable vector as first actual value reduction information.

[0066] In sub-step 3, the execution entity may determine first loss information for the first initial value reduction information and the first actual value reduction information, wherein the first loss information may represent value difference information between the first initial value reduction information and the first actual value reduction information.

[0067] As an example, the execution entity may generate first loss information for the first initial value reduction information and the first actual value reduction information using a cross entropy loss function.

[0068] Sub-step 4: in response to determining that the first loss information is greater than a first value, the execution entity may determine the first initial value reduction information generation model as a first value reduction information generation model, wherein the first value may be a preset value.

[0069] In the second step, in response to determining that the above-mentioned first loss information is less than or equal to the above-mentioned first numerical value, according to the above-mentioned first loss information, the above-mentioned execution entity can perform model parameter training on the above-mentioned first initial value reduction information generation model to obtain an updated first value reduction information generation model, and re-select the first initial adjustment variable vector from the above-mentioned initial adjustment variable vector set, and determine the above-mentioned updated first value reduction information generation model as the first initial value reduction information generation model to continue to execute the above-mentioned first training step.

[0070] In some optional implementations of some embodiments, the above-mentioned initial discriminant model set included in the initial discriminant model set of the initial target value payment method probability generation model is trained according to the above-mentioned initial variable vector set, the actual value reduction information set and the actual value payment method information set to obtain the discriminant model set, which may include the following steps:

[0071] In the first step, the execution subject may select an initial confusion variable vector from the initial confusion variable vector set as the first initial confusion variable vector, and perform the following second training step:

[0072] In sub-step 1, the execution entity may determine the actual value reduction information corresponding to the first initial confounding variable vector as the second actual value reduction information, and determine the initial instrumental variable vector corresponding to the first initial confounding variable vector as the first initial instrumental variable vector.

[0073] Sub-step 2: the execution entity may generate a first actual value reduction vector for the second actual value reduction information.

[0074] As an example, the execution entity may use the initial feature data vector representation model to generate a first actual value reduction vector for the second actual value reduction information.

[0075] Sub-step 3, inputting the first initial confounding variable vector and the first actual value reduction vector into the first initial value payment method probability generation model to generate the first initial value payment method probability information. The first initial value payment method probability generation model may be a first value payment method probability generation model whose model training has not yet been completed. The first value payment method probability generation model may be a model that generates value payment method prediction probability information based on confounding variable data and value reduction information.

[0076] Sub-step 4, inputting the first initial instrumental variable vector, the first initial confounding variable vector and the first actual value reduction vector into the second initial value payment method probability generation model to generate the second initial value payment method probability information. The second initial value payment method probability generation model may be a second value payment method probability generation model whose model training has not yet been completed. The second value payment method probability generation model may be a model that generates value payment method prediction probability information based on instrumental variable data, confounding variable data and value reduction information.

[0077] Sub-step 5: Generate second loss information for the first initial value payment method probability information and the second initial value payment method probability information, wherein the second loss information may represent the probability information difference between the first initial value payment method probability information and the second initial value payment method probability information.

[0078] As an example, the execution entity may generate second loss information for the first initial value payment method probability information and the second initial value payment method probability information using a cross entropy loss function.

[0079] Sub-step 6, in response to determining that the second loss information is greater than a second value, determining the first initial value payment method probability generation model as the first value payment method probability generation model, and determining the second initial value payment method probability generation model as the second value payment method probability generation model. The second value may be a number of a predetermined value.

[0080] The second step, in response to determining that the above-mentioned second loss information is less than or equal to the above-mentioned second numerical value, according to the above-mentioned second loss information, the model parameters of the above-mentioned first initial value payment method probability generation model and the second initial value payment method probability generation model are trained to obtain the updated first value payment method probability generation model and the updated first value payment method probability generation model, and re-select the first initial confusion variable vector from the above-mentioned initial confusion variable vector set, use the above-mentioned updated first value payment method probability generation model as the first initial value payment method probability generation model, and use the above-mentioned updated second value payment method probability generation model as the second initial value payment method probability generation model, so as to continue to execute the above-mentioned second training step.

[0081] Step 304, fix the model parameters of the above-mentioned first value reduction information generation model, the above-mentioned first value payment method probability generation model and the above-mentioned second value payment method probability generation model, and according to the above-mentioned initial variable vector set, the above-mentioned actual value reduction information set and the above-mentioned actual value payment method information set, perform model training on the initial generative model set and the above-mentioned initial feature data vector representation model included in the trained target value payment method probability generation model to obtain the target value payment method probability generation model.

[0082] In some embodiments, the execution subject may fix the model parameters of the first value reduction information generation model, the first value payment method probability generation model, and the second value payment method probability generation model, and perform model training on the initial generative model set and the initial feature data vector representation model included in the trained target value payment method probability generation model according to the initial variable vector set, the actual value reduction information set, and the actual value payment method information set, to obtain the target value payment method probability generation model. Among them, the generative model set in the target value payment method probability generation model includes: the third value payment method probability generation model, the fourth value payment method probability generation model, and the second value reduction information generation model. The model parameter fixing may be that in the training process of the target value payment method probability generation model after subsequent training, the model parameters of the first value reduction information generation model, the first value payment method probability generation model, and the second value payment method probability generation model do not change. Among them, the generative model set in the target value payment method probability generation model includes: the third value payment method probability generation model, the fourth value payment method probability generation model, and the second value reduction information generation model. The third value payment method probability generation model can be a model that generates probability information of using the target value payment method based on the third initial variable vector and the corresponding actual value reduction information. In practice, the third value payment method probability generation model can be a multi-layer serially connected recurrent neural network model. The vector content of the first initial variable vector corresponding to the first value payment method probability generation model, the vector content of the first initial variable vector corresponding to the second value payment method probability generation model, and the vector content of the first initial variable vector corresponding to the third value payment method probability generation model are different. Correspondingly, the network structure of the first value payment method probability generation model can also be different from the network structure of the second value payment method probability generation model and the network structure of the third value payment method probability generation model. Specifically, the number of network layers corresponding to the three is different. The fourth value payment method probability generation model can be a model that generates probability information of using the target value payment method based on the fourth initial variable vector. Specifically, the fourth initial variable vector is different from the first initial variable vector, the second initial variable vector, and the third initial variable vector. The network structure of the first value payment method probability generation model, the network structure of the second value payment method probability generation model, the network structure of the third value payment method probability generation model, and the network structure of the fourth value payment method probability generation model can be different.

[0083] In some optional implementations of some embodiments, the initial generative model set and the initial feature data vector representation model included in the trained target value payment method probability generation model are trained according to the initial variable vector set, the actual value reduction information set, and the actual value payment method information set to obtain the target value payment method probability generation model, which may include the following steps:

[0084] In the first step, an initial confusion variable vector is selected from the above initial confusion variable vector set as the second initial confusion variable vector, and the following third training step is performed:

[0085] Sub-step 1, determine the initial adjustment variable vector, initial instrumental variable vector, corresponding actual value reduction vector and corresponding actual value reduction information corresponding to the second initial confounding variable vector, as the second initial adjustment variable vector, the second initial instrumental variable vector, the third actual value reduction vector and the third actual value reduction information, respectively.

[0086] Sub-step 2, inputting the second initial confounding variable vector, the third actual value reduction vector and the second initial adjustment variable vector into the third initial value payment method probability generation model to generate the third initial value payment method probability information. The third initial value payment method probability generation model may be a third value payment method probability generation model whose model training has not yet been completed. The third value payment method probability generation model may be a model that generates value payment method prediction probability information based on confounding variable data, value reduction data and adjustment variable data.

[0087] Sub-step 3, input the second initial adjustment variable vector into the fourth initial value payment method probability generation model to generate the fourth initial value payment method probability information. The fourth initial value payment method probability generation model may be a fourth value payment method probability generation model whose model training has not yet been completed. The fourth value payment method probability generation model may be a model that generates value payment method prediction probability information based on adjustment variable data.

[0088] Sub-step 4: Input the second initial instrumental variable vector into the second initial value reduction information generation model to generate the second initial value reduction information. The second initial value reduction information generation model may be a second value reduction information generation model whose training has not yet been completed. The second value reduction information generation model may be a model that generates value reduction information based on instrumental variable data.

[0089] Sub-step 5: Generate third loss information for the third initial value payment method probability information and the corresponding actual value payment method information. The third loss information can represent the information difference between the third initial value payment method probability information and the corresponding actual value payment method information. For specific implementation methods, see the generation of the second loss information.

[0090] Sub-step 6, generating fourth loss information for the fourth initial value payment method probability information and the corresponding actual value payment method information. The fourth loss information can represent the information difference between the fourth initial value payment method probability information and the corresponding actual value payment method information. For specific implementation methods, see the generation of the second loss information.

[0091] Sub-step 7: Generate fifth loss information corresponding to the second initial value reduction information and the third actual value reduction information. The fifth loss information may represent the information difference between the second initial value reduction information and the third actual value reduction information. For specific implementation methods, see the generation of the second loss information.

[0092] Sub-step 8, in response to determining that the third loss information is less than the third value, the fourth loss information is less than the fourth value, and the fifth loss information is less than the fifth value, the third initial value payment method probability generation model, the fourth initial value payment method probability generation model, the second initial value reduction information generation model, and the initial feature data vector representation model are respectively determined as the third value payment method probability generation model, the fourth value payment method probability generation model, the second value reduction information generation model, and the feature data vector representation model. The third value, the fourth value, and the fifth value may be pre-set values.

[0093] Optionally, the steps further include:

[0094] The first step, in response to determining that there is at least one of the third loss information being greater than or equal to the third numerical value, the fourth loss information being greater than or equal to the fourth numerical value, and the fifth loss information being greater than or equal to the fifth numerical value, the third initial value payment method probability generation model, the fourth initial value payment method probability generation model, the second initial value reduction information generation model, and the initial feature data vector representation model are updated with model parameters according to the third loss information, the fourth loss information, and the fifth loss information, to obtain the updated third initial value payment method probability generation model, the updated fourth initial value payment method probability generation model, the updated second initial value reduction information generation model, and the updated initial feature data vector representation model, which are used as the third initial value payment method probability generation model, the fourth initial value payment method probability generation model, the second initial value reduction information generation model, and the initial feature data vector representation model, respectively, and the second initial confusion variable vector is reselected to continue to execute the third training step.

[0095] In some optional implementations of some embodiments, the target value payment method probability generation model is verified by the following steps:

[0096] The first step is to obtain a user information set. The user information may be identity information that characterizes the identity of the user. In practice, the user information may be user identification information. The user information in the user information set may be user information used to subsequently verify the model effect of the target value payment method probability generation model.

[0097] In the second step, for each of the multiple value reduction values ​​included in the value reduction information, the following information generation step is performed:

[0098] Sub-step 1, determine the value distribution value corresponding to each user information in the above user information set. Among them, the value reduction information can be the value reduction transformation information of the corresponding item. For e-commerce scenarios, the value reduction information can be a payment discount coupon. The value reduction value can be the degree of reduction transformation of the value information. Each value reduction value in the multiple value reduction values ​​is different. Specifically, for the value reduction information being a payment discount coupon, the corresponding multiple value reductions can be {"0 yuan", "1 yuan", "3 yuan", "5 yuan", "10 yuan"}. The value distribution value can be the preferential value corresponding to the payment discount coupon actually issued to the user corresponding to the user information.

[0099] Sub-step 2: Filter out a user information subset whose corresponding value distribution value is a target value from the above user information set as the first user information subset, wherein the target value may be "0".

[0100] Sub-step 3: Filter out a user information subset whose corresponding value issuance value is the above-mentioned value reduction value from the above-mentioned user information set as the second user information subset.

[0101] Sub-step 4: merge the first user information subset and the second user information subset to obtain a merged user information set.

[0102] Sub-step 5, for each fused user information in the above-mentioned fused user information set, use the above-mentioned target value payment method probability generation model to determine the first target value payment method prediction probability information of the above-mentioned fused user information under the above-mentioned value reduction value and the second target value payment method prediction probability information of the above-mentioned fused user information under the above-mentioned target value. For example, the value reduction value is "3". The first target value payment method prediction probability information of the fused user information under the above-mentioned value reduction value can be the probability information that the user corresponding to the fused user information uses the target value payment method to pay under the condition that the payment coupon corresponding to the payment discount value is 3 yuan. The second target value payment method prediction probability information of the fused user information under the above-mentioned target value can be the probability information that the user corresponding to the fused user information uses the target value payment method to pay under the condition that the payment coupon corresponding to the payment discount value is 0 yuan.

[0103] Sub-step 6, grouping the user information of each fused user information in the above fused user information set according to the information difference between the first target value payment method prediction probability information and the second target value payment method prediction probability information corresponding to each fused user information, to obtain at least one user information group.

[0104] As an example, first, for each fused user information, the corresponding information difference value is determined. Then, the number of user information groups is determined. Finally, according to the size of the information difference value, each fused user information is grouped to obtain at least one user information group.

[0105] Sub-step 7: for each user information group in the at least one user information group, determine the predicted average gain value corresponding to the user information group according to the information difference value set corresponding to the user information group.

[0106] As an example, the information difference value set corresponding to the user information group is averaged to obtain an average information difference value as the predicted average gain value corresponding to the above user information group.

[0107] Sub-step 8, determining the actual average gain value for each user information group, and obtaining at least one actual average gain value. The actual average gain value may represent the influence of the payment reduction information whose value distribution value is the value reduction value on each user information in the user information group. Specifically, the actual average gain value may be the influence degree of the payment reduction information whose value distribution value is the value reduction value on the user information to select the target value payment method to make value payment.

[0108] As an example, for each user information group, first, the user information corresponding to the target value of the value distribution value in the above user information group is determined as the first user information subgroup. Then, the user information corresponding to the value distribution value of each user information group is the above value reduction value is determined as the second user information subgroup. Finally, the proportion of the number of users corresponding to the second user information subgroup is divided by the proportion of the number of users corresponding to the first user information subgroup to obtain the actual average gain value.

[0109] Sub-step 9, compare the obtained at least one predicted average gain value with the at least one actual average gain value to obtain comparison information. There is a one-to-one correspondence between the predicted average gain value in the at least one predicted average gain value and the actual average gain value in the at least one actual average gain value. The comparison information can characterize the difference relationship between the at least one actual average gain value and the at least one predicted average gain value.

[0110] The third step is to generate a model verification result for the probability generation model of the target value payment method based on the multiple comparison information obtained.

[0111] As an example, the execution subject may generate accuracy information based on multiple comparison information, and then generate a model verification result of whether the target value payment method probability generation model has been verified based on the accuracy information.

[0112] The above-mentioned embodiments of the present disclosure have the following beneficial effects: by using the trained target value payment method probability generation model through the model training method of some embodiments of the present disclosure, the target value payment method prediction probability information can be accurately generated. Specifically, the reason why the relevant target value payment method prediction probability information is not accurate enough is that the problem of inaccurate representation of individual causal effects leads to inaccurate generated target value payment method prediction probability information. Based on this, the model training method of some embodiments of the present disclosure first obtains a training data set, wherein the training data in the above training data set includes: variable data, actual value reduction information and actual value payment method information. Here, the obtained variable data, real-time value reduction information and real-time value payment method information are used as input data and data labels for model training in the future. Then, the variable data set is input into the initial feature data vector representation model included in the initial target value payment method probability generation model to output the initial variable vector set. Here, the variable data in the variable data set can be converted into a vector form through the initial feature data vector representation model, so as to facilitate the subsequent input into the initial target value payment method probability generation model. Next, according to the above-mentioned initial variable vector set, actual value reduction information set and actual value payment method information set, each initial discriminant model in the initial discriminant model set included in the above-mentioned initial target value payment method probability generation model is trained to obtain a discriminant model set, wherein the above-mentioned discriminant model set includes: a first value reduction information generation model, a first value payment method probability generation model and a second value payment method probability generation model. Here, by first training each discriminant model in the initial discriminant model set, an accurate discriminant model set is trained to ensure that it will serve as an auxiliary training model for the initial generative model set in the future, and help the model training of the initial generative model set. In addition, the first value reduction information generation model, the first value payment method probability generation model and the second value payment method probability generation model can be discriminant models set based on variable causal inference, which can accurately characterize individual causal effects. Therefore, by setting the first value reduction information generation model, the first value payment method probability generation model and the second value payment method probability generation model as discriminators to assist the training of the subsequent third value payment method probability generation model, the fourth value payment method probability generation model and the second value reduction information generation model, the accuracy of model training can be greatly improved.Finally, the model parameters of the first value reduction information generation model, the first value payment method probability generation model and the second value payment method probability generation model are fixed, and the initial generative model set and the initial feature data vector representation model included in the trained target value payment method probability generation model are trained according to the initial variable vector set, the actual value reduction information set and the actual value payment method information set to obtain the target value payment method probability generation model, wherein the generative model set in the target value payment method probability generation model includes: the third value payment method probability generation model, the fourth value payment method probability generation model and the second value reduction information generation model. Here, the third value payment method probability generation model, the fourth value payment method probability generation model and the second value reduction information generation model are also generative models set based on causal inference, which can accurately generate the corresponding individual causal effect information. With the assistance of model training of the discriminant model set, the target value payment method probability generation model can be accurately obtained. In summary, by first training the discriminative model set and then training the generative model set, the individual causal effects can be fully characterized, so that the trained target value payment method probability generation model can be used to accurately generate the target value payment method prediction probability information.

[0113] Continue to refer Figure 4 , shows a process 400 of some embodiments of the method for generating prediction probability information according to the present disclosure. The method for generating prediction probability information comprises the following steps:

[0114] Step 401: Obtain variable data for a target user and multiple value reduction values ​​included in value reduction information.

[0115] In some embodiments, the execution subject (e.g., electronic device) of the above-mentioned prediction probability information generation method can obtain the variable data for the target user and the multiple value reduction values ​​included in the value reduction information by wired or wireless means. Specifically, the variable data and the multiple value reduction values ​​are not explained again.

[0116] Step 402, for each of the multiple value reduction values, input the variable data and the value reduction value into a pre-trained target value payment method probability generation model to generate target value payment method prediction probability information for the value reduction value.

[0117] In some embodiments, the execution subject may input the variable data and the value reduction value into a pre-trained target value payment method probability generation model for each value reduction value among the multiple value reduction values ​​to generate target value payment method prediction probability information for the value reduction value. The target value payment method probability generation model is generated based on a model training method. The target value payment method prediction probability information may be the probability information that the target user selects the target value payment method to make a value payment, under the premise that the payment discount value corresponding to the payment discount coupon is the value reduction value.

[0118] Step 403 , select the value reduction value whose corresponding target value payment method prediction probability information meets the preset information condition from the above multiple value reduction values ​​as the target value reduction value.

[0119] In some embodiments, the execution subject may select a value reduction value whose corresponding target value payment method prediction probability information satisfies a preset information condition from the multiple value reduction values ​​as the target value reduction value. The preset information condition may be the value with the largest value among the multiple value reduction values.

[0120] Step 404: Push the reduction information corresponding to the target value reduction value to the user terminal corresponding to the target user.

[0121] In some embodiments, the execution entity may push the reduction information corresponding to the target value reduction value to the user terminal corresponding to the target user, wherein the reduction information corresponding to the target value reduction value may be a payment reduction coupon with a payment reduction value equal to the target value reduction value.

[0122] The above-mentioned various embodiments of the present disclosure have the following beneficial effects: accurate reduction information can be pushed to target users through the prediction probability information method of some embodiments of the present disclosure.

[0123] Further references Figure 5 , shows a process 500 of another embodiment of the method for generating prediction probability information according to the present disclosure. The method for generating prediction probability information comprises the following steps:

[0124] Step 501: Obtain variable data for a target user and multiple value reduction values ​​included in value reduction information.

[0125] In some embodiments, the execution subject (eg, electronic device) may obtain the variable data for the target user and the multiple value reduction values ​​included in the value reduction information through a wired or wireless method.

[0126] Step 502, for each of the above-mentioned multiple value reduction values, the above-mentioned variable data and the above-mentioned value reduction value are input into the third value payment method probability generation model included in the target value payment method probability generation model to generate the target value payment method prediction probability information for the above-mentioned value reduction value.

[0127] In some embodiments, the above-mentioned execution entity can input the above-mentioned variable data and the above-mentioned value reduction value into the third value payment method probability generation model included in the target value payment method probability generation model for each value reduction value among the above-mentioned multiple value reduction values, so as to generate the target value payment method prediction probability information for the above-mentioned value reduction value.

[0128] Step 503 , select the value reduction value whose corresponding target value payment method prediction probability information meets the preset information condition from the above multiple value reduction values ​​as the target value reduction value.

[0129] Step 504: Push the reduction information corresponding to the target value reduction value to the user terminal corresponding to the target user.

[0130] In some embodiments, the specific implementation of steps 501, 503-504 and the technical effects thereof can be referred to in Figure 4 The corresponding steps 401, 403-404 in the embodiment will not be described in detail here.

[0131] from Figure 5 It can be seen that Figure 4 Compared with the description of some corresponding embodiments, Figure 5 The process 500 of the prediction probability information generation method in some corresponding embodiments can accurately generate the target value payment method prediction probability information for the value reduction value by using the third value payment method probability generation model.

[0132] Further references Figure 6 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a model training device, and these device embodiments are Figure 3 Corresponding to the method embodiments shown, the model training device can be specifically applied to various electronic devices.

[0133] like Figure 6As shown, a model training device 600 includes: a first acquisition unit 601, a first generation unit 602, a first training unit 603 and a second training unit 604. The first acquisition unit 601 is configured to acquire a training data set, wherein the training data in the training data set includes: variable data, actual value reduction information and actual value payment method information; the first generation unit 602 is configured to input the variable data set into the initial feature data vector representation model included in the initial target value payment method probability generation model to output an initial variable vector set; the first training unit 603 is configured to perform model training on each initial discriminant model set included in the initial target value payment method probability generation model according to the initial variable vector set, the actual value reduction information set and the actual value payment method information set, to obtain a discriminant model set, wherein the discriminant model set includes: the first value reduction information generation model, the first value payment method conception model, the first discriminant model set ... rate generation model and a second value payment method probability generation model; the second training unit 604 is configured to fix the model parameters of the above-mentioned first value reduction information generation model, the above-mentioned first value payment method probability generation model and the above-mentioned second value payment method probability generation model, and according to the above-mentioned initial variable vector set, the above-mentioned actual value reduction information set and the above-mentioned actual value payment method information set, the initial generative model set included in the trained target value payment method probability generation model and the above-mentioned initial feature data vector representation model are trained to obtain the target value payment method probability generation model, wherein the generative model set in the above-mentioned target value payment method probability generation model includes: a third value payment method probability generation model, a fourth value payment method probability generation model, and a second value reduction information generation model.

[0134] In some optional implementations of some embodiments, the variable data in the above-mentioned variable data set include: instrumental variable data, adjustment variable data, and confusion variable data; and the first generation unit 602 can be further configured to: input the instrumental variable data set, the adjustment variable data set, and the confusion variable data set into the above-mentioned initial feature data vector representation model to generate an initial instrumental variable vector set, an initial adjustment variable vector set, and an initial confusion variable vector set; and perform vector set fusion on the above-mentioned initial instrumental variable vector set, the above-mentioned initial adjustment variable vector set, and the above-mentioned initial confusion variable vector set to obtain an initial variable vector set.

[0135] In some optional implementations of some embodiments, the first training unit 603 may be further configured to: select an initial adjustment variable vector from the above-mentioned initial adjustment variable vector set as the first initial adjustment variable vector, and perform the following first training step: input the above-mentioned first initial adjustment variable vector into the first initial value reduction information generation model included in the above-mentioned initial target value payment method probability generation model to generate the first initial value reduction information; determine the actual value reduction information corresponding to the above-mentioned initial adjustment variable vector as the first actual value reduction information; determine the first initial value reduction information and the first actual value reduction information. in response to determining that the first loss information is greater than the first numerical value, determining the first initial value reduction information generation model as the first value reduction information generation model; in response to determining that the first loss information is less than or equal to the first numerical value, updating the model parameters of the first initial value reduction information generation model according to the first loss information to obtain an updated first value reduction information generation model, and reselecting the first initial adjustment variable vector from the initial adjustment variable vector set, determining the updated first value reduction information generation model as the first initial value reduction information generation model, so as to continue to execute the first training step.

[0136] In some optional implementations of some embodiments, the first training unit 603 may be further configured to: select an initial confounding variable vector from the above-mentioned initial confounding variable vector set as the first initial confounding variable vector, and perform the following second training step: determine the actual value reduction information corresponding to the above-mentioned first initial confounding variable vector as the second actual value reduction information, and determine the initial instrumental variable vector corresponding to the above-mentioned first initial confounding variable vector as the first initial instrumental variable vector; generate a first actual value reduction vector for the above-mentioned second actual value reduction information; input the above-mentioned first initial confounding variable vector and the above-mentioned first actual value reduction vector into the first initial value payment method probability generation model to generate the first initial value payment method probability information; input the above-mentioned first initial instrumental variable vector, the above-mentioned first initial confounding variable vector and the above-mentioned first actual value reduction vector into the second initial value payment method probability generation model to generate the second initial value payment method probability information; generate a first initial value payment method probability information for the above-mentioned first initial value payment method probability information and the above-mentioned the second loss information of the second initial value payment method probability information; in response to determining that the above-mentioned second loss information is greater than the second numerical value, determining the above-mentioned first initial value payment method probability generation model as the first value payment method probability generation model, and determining the above-mentioned second initial value payment method probability generation model as the second value payment method probability generation model; in response to determining that the above-mentioned second loss information is less than or equal to the above-mentioned second numerical value, according to the above-mentioned second loss information, training the model parameters of the above-mentioned first initial value payment method probability generation model and the second initial value payment method probability generation model to obtain the updated first value payment method probability generation model and the updated first value payment method probability generation model, and re-selecting the first initial confusion variable vector from the above-mentioned initial confusion variable vector set, using the above-mentioned updated first value payment method probability generation model as the first initial value payment method probability generation model, and using the above-mentioned updated second value payment method probability generation model as the second initial value payment method probability generation model, so as to continue to execute the above-mentioned second training step.

[0137] In some optional implementations of some embodiments, the second training unit 604 may be further configured to: select an initial confounding variable vector from the above-mentioned initial confounding variable vector set as the second initial confounding variable vector, and perform the following third training step: determine the initial adjustment variable vector, the initial instrumental variable vector, the corresponding actual value reduction vector and the corresponding actual value reduction information corresponding to the above-mentioned second initial confounding variable vector, as the second initial adjustment variable vector, the second initial instrumental variable vector, the third actual value reduction vector and the third actual value reduction information respectively; input the above-mentioned second initial confounding variable vector, the above-mentioned third actual value reduction vector and the above-mentioned second initial adjustment variable vector into the third initial value payment method probability generation model to generate the third initial value payment method probability information; input the above-mentioned second initial adjustment variable vector into the fourth initial value payment method probability generation model to generate the fourth initial value payment method probability information; input the above-mentioned second initial instrumental ... The amount is input into the second initial value reduction information generation model to generate the second initial value reduction information; generate the third loss information for the above-mentioned third initial value payment method probability information and the corresponding actual value payment method information; generate the fourth loss information for the above-mentioned fourth initial value payment method probability information and the corresponding actual value payment method information; generate the fifth loss information corresponding to the above-mentioned second initial value reduction information and the above-mentioned third actual value reduction information; in response to determining that the above-mentioned third loss information is less than the third numerical value, the above-mentioned fourth loss information is less than the fourth numerical value, and the above-mentioned fifth loss information is less than the fifth numerical value, the above-mentioned third initial value payment method probability generation model, the above-mentioned fourth initial value payment method probability generation model, the above-mentioned second initial value reduction information generation model and the above-mentioned initial feature data vector representation model are respectively determined as the third value payment method probability generation model, the fourth value payment method probability generation model, the second value reduction information generation model and the feature data vector representation model.

[0138] It is understandable that the units recorded in the model training device 600 are similar to the reference Figure 3 Therefore, the operations, features and beneficial effects described above for the method are also applicable to the model training device 600 and the units contained therein, and will not be repeated here.

[0139] Further references Figure 7 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a prediction probability information generating device, and these device embodiments are Figure 4 Corresponding to the method embodiments shown, the prediction probability information generating device can be specifically applied to various electronic devices.

[0140] like Figure 7As shown, a prediction probability information generating device 700 includes: a second acquisition unit 701, a second generation unit 702, a screening unit 703 and a push unit 704. The second acquisition unit 701 is configured to acquire variable data for a target user and multiple value reduction values ​​included in the value reduction information; the second generation unit 702 is configured to input the variable data and the value reduction value into a pre-trained target value payment method probability generation model for each value reduction value in the multiple value reduction values, so as to generate target value payment method prediction probability information for the value reduction value, wherein the target value payment method probability generation model is generated based on a model training method; the screening unit 703 is configured to screen out the value reduction value corresponding to the target value payment method prediction probability information that meets the preset information condition from the multiple value reduction values ​​as the target value reduction value; the push unit 704 is configured to push the reduction information corresponding to the target value reduction value to the user terminal corresponding to the target user.

[0141] In some optional implementations of some embodiments, the second generation unit 702 may be further configured to: for each value reduction value among the above-mentioned multiple value reduction values, input the above-mentioned variable data and the above-mentioned value reduction value into the third value payment method probability generation model included in the target value payment method probability generation model, so as to generate the target value payment method predicted probability information for the above-mentioned value reduction value.

[0142] It can be understood that the units recorded in the prediction probability information generating device 700 are similar to those in the reference Figure 4 Therefore, the operations, features and beneficial effects described above for the method are also applicable to the prediction probability information generating device 700 and the units included therein, and will not be described in detail here.

[0143] Reference below Figure 8 , which shows an electronic device (eg, Figure 1 Schematic diagram of the structure of the electronic device 101)800. Figure 8 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0144] like Figure 8As shown, the electronic device 800 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 801, which can perform various appropriate actions and processes according to a program stored in a read-only memory 802 or a program loaded from a storage device 808 into a random access memory 803. Various programs and data required for the operation of the electronic device 800 are also stored in the random access memory 803. The processing device 801, the read-only memory 802, and the random access memory 803 are connected to each other via a bus 804. An input / output interface 805 is also connected to the bus 804.

[0145] Typically, the following devices may be connected to the input / output interface 805: an input device 806 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 807 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 808 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 809. The communication device 809 may allow the electronic device 800 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 8 The electronic device 800 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 8 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0146] In particular, according to some embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from a network through a communication device 809, or installed from a storage device 808, or installed from a read-only memory 802. When the computer program is executed by the processing device 801, the above-mentioned functions defined in the method of some embodiments of the present disclosure are executed.

[0147] It should be noted that the computer-readable medium in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, 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 above. In some embodiments of the present disclosure, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In some embodiments of the present disclosure, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0148] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with 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"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0149] The computer-readable medium may be included in the electronic device, or may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device: obtains a training data set, wherein the training data in the training data set includes: variable data, actual value reduction information, and actual value payment method information;

[0150] Inputting the variable data set into the initial feature data vector representation model included in the initial target value payment method probability generation model to output an initial variable vector set;

[0151] According to the above-mentioned initial variable vector set, the actual value reduction information set and the actual value payment method information set, model training is performed on each initial discriminant model set included in the above-mentioned initial target value payment method probability generation model to obtain a discriminant model set, wherein the above-mentioned discriminant model set includes: a first value reduction information generation model, a first value payment method probability generation model and a second value payment method probability generation model; the model parameters of the above-mentioned first value reduction information generation model, the above-mentioned first value payment method probability generation model and the above-mentioned second value payment method probability generation model are fixed, and according to the above-mentioned initial variable vector set, the above-mentioned actual value reduction information set and the above-mentioned actual value payment method information set, model training is performed on the initial generative model set and the above-mentioned initial feature data vector representation model included in the trained target value payment method probability generation model to obtain a target value payment method probability generation model, wherein the generative model set in the above-mentioned target value payment method probability generation model includes: a third value payment method probability generation model, a fourth value payment method probability generation model and a second value reduction information generation model. Acquire multiple value reduction values ​​including variable data and value reduction information for a target user; for each value reduction value among the above multiple value reduction values, input the above variable data and the above value reduction value into a pre-trained target value payment method probability generation model to generate target value payment method prediction probability information for the above value reduction value, wherein the above target value payment method probability generation model is generated based on a model training method; select the value reduction value whose corresponding target value payment method prediction probability information meets the preset information conditions from the above multiple value reduction values ​​as the target value reduction value; push the reduction information corresponding to the above target value reduction value to the user terminal corresponding to the above target user.

[0152] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0153] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0154] The units described in some embodiments of the present disclosure may be implemented by software or by hardware. The described units may also be provided in a processor, for example, may be described as: a processor includes a first acquisition unit, a first generation unit, a first training unit, and a second training unit. The names of these units do not, in some cases, constitute limitations on the units themselves, for example, the first acquisition unit may also be described as a "unit for acquiring a training data set".

[0155] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0156] Some embodiments of the present disclosure also provide a computer program product, including a computer program, which implements any of the above-mentioned model training methods or prediction probability information generation methods when executed by a processor.

[0157] The above descriptions are only some preferred embodiments of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with the technical features with similar functions disclosed in the embodiments of the present disclosure (but not limited to) and the technical solutions formed.

Claims

1. A model training method, comprising: Acquire a training data set, wherein the training data in the training data set includes: variable data, actual value reduction information and actual value payment method information; Inputting the variable data set into the initial feature data vector representation model included in the initial target value payment method probability generation model to output an initial variable vector set; According to the initial variable vector set, the actual value reduction information set and the actual value payment method information set, each initial discriminant model in the initial discriminant model set included in the initial target value payment method probability generation model is trained to obtain a discriminant model set, wherein the discriminant model set includes: a first value reduction information generation model, a first value payment method probability generation model and a second value payment method probability generation model; The model parameters of the first value reduction information generation model, the first value payment method probability generation model and the second value payment method probability generation model are fixed, and the initial generative model set and the initial feature data vector representation model included in the trained target value payment method probability generation model are trained according to the initial variable vector set, the actual value reduction information set and the actual value payment method information set to obtain the target value payment method probability generation model, wherein the generative model set in the target value payment method probability generation model includes: a third value payment method probability generation model, a fourth value payment method probability generation model and a second value reduction information generation model.

2. The method according to claim 1, wherein: The variable data in the variable data set include: instrumental variable data, adjustment variable data, and confounding variable data; and The step of inputting the variable data set into the initial feature data vector representation model included in the initial target value payment method probability generation model to output an initial variable vector set includes: Inputting the instrumental variable data set, the adjustment variable data set and the confounding variable data set into the initial feature data vector characterization model to generate an initial instrumental variable vector set, an initial adjustment variable vector set and an initial confounding variable vector set; The initial instrumental variable vector set, the initial adjustment variable vector set and the initial confounding variable vector set are fused to obtain an initial variable vector set.

3. The method according to claim 2, wherein: The method of training each initial discriminant model in the initial discriminant model set included in the initial target value payment method probability generation model according to the initial variable vector set, the actual value reduction information set and the actual value payment method information set, to obtain a discriminant model set, including: An initial adjustment variable vector is selected from the initial adjustment variable vector set as a first initial adjustment variable vector, and the following first training step is performed: Inputting the first initial adjustment variable vector into the first initial value reduction information generation model included in the initial target value payment method probability generation model to generate first initial value reduction information; determining actual value reduction information corresponding to the first initial adjustment variable vector as first actual value reduction information; determining first loss information for the first initial value reduction information and the first actual value reduction information; In response to determining that the first loss information is greater than a first value, determining the first initial value reduction information generation model as a first value reduction information generation model; In response to determining that the first loss information is less than or equal to the first numerical value, the model parameters of the first initial value reduction information generation model are updated according to the first loss information to obtain an updated first value reduction information generation model, and the first initial adjustment variable vector is reselected from the initial adjustment variable vector set, and the updated first value reduction information generation model is determined as the first initial value reduction information generation model to continue executing the first training step.

4. The method according to claim 2, wherein: The method of training each initial discriminant model in the initial discriminant model set included in the initial target value payment method probability generation model according to the initial variable vector set, the actual value reduction information set and the actual value payment method information set, to obtain a discriminant model set, including: An initial confusion variable vector is selected from the initial confusion variable vector set as a first initial confusion variable vector, and the following second training step is performed: Determine the actual value reduction information corresponding to the first initial confounding variable vector as the second actual value reduction information, and determine the initial instrumental variable vector corresponding to the first initial confounding variable vector as the first initial instrumental variable vector; generating a first actual value reduction vector for the second actual value reduction information; Inputting the first initial confounding variable vector and the first actual value reduction vector into a first initial value payment method probability generation model to generate first initial value payment method probability information; Inputting the first initial instrumental variable vector, the first initial confounding variable vector and the first actual value reduction vector into a second initial value payment method probability generation model to generate second initial value payment method probability information; generating second loss information for the first initial value payment method probability information and the second initial value payment method probability information; In response to determining that the second loss information is greater than a second value, determining the first initial value payment method probability generation model as a first value payment method probability generation model, and determining the second initial value payment method probability generation model as a second value payment method probability generation model; In response to determining that the second loss information is less than or equal to the second numerical value, the model parameters of the first initial value payment method probability generation model and the second initial value payment method probability generation model are updated according to the second loss information to obtain an updated first value payment method probability generation model and an updated first value payment method probability generation model, and the first initial confusion variable vector is re-selected from the initial confusion variable vector set, and the updated first value payment method probability generation model is used as the first initial value payment method probability generation model, and the updated second value payment method probability generation model is used as the second initial value payment method probability generation model to continue to execute the second training step.

5. The method according to claim 2, wherein: The method of performing model training on the initial generative model set and the initial feature data vector representation model included in the trained target value payment method probability generation model according to the initial variable vector set, the actual value reduction information set and the actual value payment method information set, to obtain the target value payment method probability generation model, including: An initial confusion variable vector is selected from the initial confusion variable vector set as a second initial confusion variable vector, and the following third training step is performed: Determine an initial adjustment variable vector, an initial instrumental variable vector, a corresponding actual value reduction vector, and a corresponding actual value reduction information corresponding to the second initial confounding variable vector, as a second initial adjustment variable vector, a second initial instrumental variable vector, a third actual value reduction vector, and a third actual value reduction information, respectively; Inputting the second initial confounding variable vector, the third actual value reduction vector and the second initial adjustment variable vector into a third initial value payment method probability generation model to generate third initial value payment method probability information; Inputting the second initial adjustment variable vector into a fourth initial value payment method probability generation model to generate fourth initial value payment method probability information; Inputting the second initial instrumental variable vector into a second initial value impairment information generation model to generate second initial value impairment information; Generate third loss information for the third initial value payment method probability information and the corresponding actual value payment method information; Generate fourth loss information for the fourth initial value payment method probability information and the corresponding actual value payment method information; generating fifth loss information corresponding to the second initial value reduction information and the third actual value reduction information; In response to determining that the third loss information is less than the third numerical value, the fourth loss information is less than the fourth numerical value, and the fifth loss information is less than the fifth numerical value, the third initial value payment method probability generation model, the fourth initial value payment method probability generation model, the second initial value reduction information generation model, and the initial feature data vector representation model are respectively determined as the third value payment method probability generation model, the fourth value payment method probability generation model, the second value reduction information generation model, and the feature data vector representation model.

6. The method according to claim 1, wherein: The target value payment method probability generation model is verified by the following steps: Get user information set; For each of the multiple value reduction values ​​included in the value reduction information, the following information generation steps are performed: Determine the value distribution value corresponding to each user information in the user information set; Filtering out a user information subset whose corresponding value distribution value is a target value from the user information set as a first user information subset; Filtering out a user information subset corresponding to the value issuance value and the value reduction value from the user information set as a second user information subset; Merging the first user information subset and the second user information subset to obtain a merged user information set; For each fused user information in the fused user information set, using the target value payment method probability generation model, determine the first target value payment method prediction probability information of the fused user information at the value reduction value and the second target value payment method prediction probability information of the fused user information at the target value; According to the information difference between the first target value payment method prediction probability information and the second target value payment method prediction probability information corresponding to each fused user information, each fused user information in the fused user information set is grouped to obtain at least one user information group; For each user information group in the at least one user information group, determining a predicted average gain value corresponding to the user information group according to an information difference value set corresponding to the user information group; Determine an actual average gain value for each user information group to obtain at least one actual average gain value; Comparing the obtained at least one predicted average gain value with the at least one actual average gain value to obtain comparison information; Based on the obtained multiple comparison information, a model verification result of the probability generation model for the target value payment method is generated.

7. A method for generating prediction probability information, comprising: Obtaining variable data for a target user and a plurality of value reduction values ​​included in value reduction information; For each value reduction value among the plurality of value reduction values, inputting the variable data and the value reduction value into a pre-trained target value payment method probability generation model to generate target value payment method prediction probability information for the value reduction value, wherein the target value payment method probability generation model is generated based on the method according to any one of claims 1 to 6; Filter out, from the plurality of value reduction values, a value reduction value corresponding to the target value payment method prediction probability information that satisfies the preset information condition as the target value reduction value; The reduction information corresponding to the target value reduction value is pushed to the user terminal corresponding to the target user.

8. The method according to claim 7, wherein: For each of the plurality of value reduction values, the variable data and the value reduction value are input into a pre-trained target value payment method probability generation model to generate target value payment method prediction probability information for the value reduction value, including: For each of the multiple value reduction values, the variable data and the value reduction value are input into a third value payment method probability generation model included in the target value payment method probability generation model to generate target value payment method prediction probability information for the value reduction value.

9. A model training device, comprising: A first acquisition unit is configured to acquire a training data set, wherein the training data in the training data set includes: variable data, actual value reduction information and actual value payment method information; A first generating unit is configured to input the variable data set into the initial feature data vector representation model included in the initial target value payment method probability generation model to output an initial variable vector set; The first training unit is configured to perform model training on each initial discriminant model in the initial discriminant model set included in the initial target value payment method probability generation model according to the initial variable vector set, the actual value reduction information set and the actual value payment method information set, so as to obtain a discriminant model set, wherein the discriminant model set includes: a first value reduction information generation model, a first value payment method probability generation model and a second value payment method probability generation model; The second training unit is configured to fix the model parameters of the first value reduction information generation model, the first value payment method probability generation model and the second value payment method probability generation model, and perform model training on the initial generative model set and the initial feature data vector representation model included in the trained target value payment method probability generation model according to the initial variable vector set, the actual value reduction information set and the actual value payment method information set to obtain the target value payment method probability generation model, wherein the generative model set in the target value payment method probability generation model includes: a third value payment method probability generation model, a fourth value payment method probability generation model and a second value reduction information generation model.

10. A prediction probability information generating device, comprising: A second acquisition unit is configured to acquire variable data for a target user and a plurality of value reduction values ​​included in the value reduction information; a second generating unit configured to input the variable data and the value reduction value into a pre-trained target value payment method probability generating model for each value reduction value of the plurality of value reduction values, so as to generate target value payment method prediction probability information for the value reduction value, wherein the target value payment method probability generating model is generated based on the method according to any one of claims 1 to 6; A screening unit is configured to screen out, from the plurality of value reduction values, a value reduction value corresponding to which the target value payment method prediction probability information satisfies a preset information condition as a target value reduction value; The push unit is configured to push the reduction information corresponding to the target value reduction value to the user terminal corresponding to the target user.

11. An electronic device, comprising: one or more processors; a storage device having one or more programs stored thereon, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 8.

12. A computer readable medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

13. A computer program product, comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 8.

Images