CN110046799B - Decision optimization method and device - Google Patents
Decision optimization method and device Download PDFInfo
- Publication number
- CN110046799B CN110046799B CN201910176389.8A CN201910176389A CN110046799B CN 110046799 B CN110046799 B CN 110046799B CN 201910176389 A CN201910176389 A CN 201910176389A CN 110046799 B CN110046799 B CN 110046799B
- Authority
- CN
- China
- Prior art keywords
- decision
- variable
- characteristic
- variables
- decision characteristic
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Educational Administration (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The disclosure provides a decision optimization method and a decision optimization device. The method comprises the following steps: determining the contribution degree of a prediction result of each decision characteristic variable in a prediction model under user characteristic data, wherein the prediction model is established based on a decision target and comprises the decision characteristic variable and a non-decision characteristic variable; constructing a decision characteristic variable combination to be optimized based on the determined prediction result contribution of each decision characteristic variable; optimizing the variable values of each decision characteristic variable in the constructed decision characteristic variable combination so as to ensure that the corresponding prediction result of the prediction model is optimal; and performing decision optimization processing according to the variable values of the decision characteristic variables in the decision characteristic variable combination obtained after optimization processing. By the method, personalized decision optimization aiming at the user can be realized.
Description
Technical Field
The present disclosure relates generally to the field of computer technology, and more particularly, to a method and apparatus for decision optimization.
Background
In many business scenarios, a service provider may wish to provide a customized service to a user to maximize the user's experience, to secure the user's rights, or to achieve other business goals. For example, in a marketing scenario, the registration conversion or retention rate of a customer is maximized. In transaction risk management, the risk of a user being defrauded per transaction fund is minimized. The nature of such demands is that personalized decisions are made for different users, and the personalized decisions can achieve the optimal decision effect, thereby achieving the decision goal to the maximum extent.
Disclosure of Invention
In view of the foregoing, the present disclosure provides a method and apparatus for decision optimization. By utilizing the method and the device, personalized decision optimization can be realized for different users.
According to an aspect of the present disclosure, there is provided a method for decision optimization, comprising: determining the contribution degree of a prediction result of each decision characteristic variable in a prediction model under user characteristic data, wherein the prediction model is established based on a decision target and comprises the decision characteristic variable and a non-decision characteristic variable; constructing a decision characteristic variable combination to be optimized based on the determined prediction result contribution of each decision characteristic variable; optimizing the variable values of each decision characteristic variable in the constructed decision characteristic variable combination so as to ensure that the corresponding prediction result of the prediction model is optimal; and performing decision optimization processing according to the variable values of the decision characteristic variables in the decision characteristic variable combination obtained after optimization processing.
Optionally, in an example of the above aspect, determining a prediction result contribution degree of each decision characteristic variable in the prediction model under the user characteristic data includes: and determining the prediction result contribution degree of each decision characteristic variable in the prediction model under the user characteristic data by using the interpretation model.
Optionally, in an example of the above aspect, the interpretation model comprises one of the following interpretation models: the Shap value model, the LIME model, and the DeepLift model.
Optionally, in an example of the above aspect, constructing a decision feature variable combination to be optimized based on the determined prediction result contribution degrees of the respective decision feature variables may include: based on the determined contribution degree of the prediction result of each decision characteristic variable, sequencing each decision characteristic variable; and selecting a preset number of decision characteristic variables with higher contribution degree from the sorted decision characteristic variables as a decision characteristic variable combination to be optimized.
Optionally, in an example of the above aspect, the optimizing the variable values of the decision characteristic variables in the constructed decision characteristic variable combination may include: optimizing the variable values of each decision characteristic variable in the constructed decision characteristic variable combination by using one of the following optimization algorithms: particle swarm algorithm, genetic algorithm and annealing algorithm.
Optionally, in an example of the above aspect, the optimizing the variable values of the decision characteristic variables in the constructed decision characteristic variable combination may include: and in a preset decision variable value range, optimizing the variable value of each decision characteristic variable in the constructed decision characteristic variable combination.
According to another aspect of the present disclosure, there is provided an apparatus for decision optimization, comprising: a contribution degree determination unit configured to determine a prediction result contribution degree of each decision characteristic variable in a prediction model under the user characteristic data, wherein the prediction model is created based on a decision target, and the prediction model comprises decision characteristic variables and non-decision characteristic variables; the decision characteristic variable combination construction unit is configured to construct a decision characteristic variable combination to be optimized based on the determined prediction result contribution degree of each decision characteristic variable; an optimization processing unit configured to perform optimization processing on the variable values of each decision characteristic variable in the constructed decision characteristic variable combination so as to optimize the corresponding prediction result of the prediction model; and the decision optimization unit is configured to perform decision optimization processing according to the variable values of the decision characteristic variables in the decision characteristic variable combination obtained after optimization processing.
Optionally, in an example of the above aspect, the contribution degree determining unit is configured to: and determining the prediction result contribution degree of each decision characteristic variable in the prediction model under the user characteristic data by using the interpretation model.
Optionally, in an example of the above aspect, the interpretation model comprises one of the following interpretation models: the Shap value model, the LIME model, and the DeepLift model.
Optionally, in an example of the above aspect, the decision feature variable combination construction unit includes: the sequencing module is configured to sequence the decision characteristic variables based on the determined prediction result contribution degrees of the decision characteristic variables; and the characteristic selection module is configured to select a predetermined number of decision characteristic variables with higher contribution degree from the sorted decision characteristic variables as a decision characteristic variable combination to be optimized.
Optionally, in one example of the above aspect, the optimizing processing unit is configured to: and optimizing the variable values of the decision characteristic variables in the constructed decision characteristic variable combination by using one of the following optimization algorithms: particle swarm algorithm, genetic algorithm and annealing algorithm.
Optionally, in one example of the above aspect, the optimizing processing unit is configured to: and in a preset decision variable value range, optimizing the variable value of each decision characteristic variable in the constructed decision characteristic variable combination.
According to another aspect of the present disclosure, there is provided a computing device comprising: at least one processor, and a memory coupled with the at least one processor, the memory storing instructions that, when executed by the at least one processor, cause the at least one processor to perform a method for decision optimization as described above.
According to another aspect of the present disclosure, there is provided a non-transitory machine-readable storage medium storing executable instructions that, when executed, cause the machine to perform the method for decision optimization as described above.
Drawings
A further understanding of the nature and advantages of the present disclosure may be realized by reference to the following drawings. In the drawings, similar components or features may have the same reference numerals.
FIG. 1 shows a flow diagram of a decision optimization method according to an embodiment of the present disclosure;
FIG. 2 shows a flow diagram of one example of a decision feature variable prediction result contribution determination process, according to an embodiment of the present disclosure;
FIG. 3 shows a flow diagram of one example of a decision feature variable combination construction process, in accordance with an embodiment of the present disclosure;
FIG. 4 shows a flow diagram of one example of an online switching process for a personalized decision mechanism, according to an embodiment of the disclosure;
FIG. 5 shows a block diagram of a decision optimization device according to an embodiment of the present disclosure;
FIG. 6 shows a block diagram of one example of a decision feature variable combination construction unit according to an embodiment of the present disclosure;
FIG. 7 illustrates a block diagram of a computing device for decision optimization in accordance with an embodiment of the present disclosure.
Detailed Description
The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and thereby implement the subject matter described herein, and are not intended to limit the scope, applicability, or examples set forth in the claims. Changes may be made in the function and arrangement of elements discussed without departing from the scope of the disclosure. Various examples may omit, substitute, or add various procedures or components as needed. For example, the described methods may be performed in an order different from that described, and various steps may be added, omitted, or combined. In addition, features described with respect to some examples may also be combined in other examples.
As used herein, the term "include" and its variants mean open-ended terms in the sense of "including, but not limited to. The term "based on" means "based at least in part on". The terms "one embodiment" and "an embodiment" mean "at least one embodiment". The term "another embodiment" means "at least one other embodiment". The terms "first," "second," and the like may refer to different or the same object. Other definitions, whether explicit or implicit, may be included below. The definition of a term is consistent throughout the specification unless the context clearly dictates otherwise.
Decision optimization methods and apparatuses according to embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.
Fig. 1 shows a flow diagram of a decision optimization method according to an embodiment of the present disclosure.
As shown in FIG. 1, at block 110, the prediction result contribution of each decision feature variable in the prediction model under the user feature data is determined. Here, the predictive model is created based on a decision target, and includes decision-making characteristic variables and non-decision-making characteristic variables. The user characteristic data is user characteristic data corresponding to the decision characteristic variable and the non-decision characteristic variable.
For example, assume that in a marketing scenario, the decision target is the number of users registered for conversion, and in a transaction risk control scenario, the decision target is the sum of the amounts lost by the users due to fraudulent transactions. For this decision objective, an appropriate set of characteristic variables is created to build a predictive model to predict the decision objective based on the user characteristic data, such as predicting whether the user is converting or not, or predicting the amount of money lost by fraud in the user's transaction. In the present disclosure, the term "non-decision feature variable" refers to a feature variable that a decision party cannot interfere with, such as the age, the character, the past behavior history of the user, and the like. The term "decision characteristic variable" refers to a characteristic variable that a decision party can intervene in changing, for example, a characteristic variable such as a benefit and a profit issued to a user in a case where a decision target is the number of registered conversion persons of the user in a marketing scenario, or a characteristic variable such as a risk reminder output to the user in a transaction risk control scenario.
In addition, in the present disclosure, the prediction model may be created by using training data based on a decision target in advance, or may be created in real time by using training data after the decision target is obtained. Also, the decision target may be input in advance or in real time. In addition, in order to improve the accuracy of the prediction model, the training data used for prediction model training should cover the user population with as many features as possible, and try as many decision variable values as possible, thereby ensuring the richness of the training data dimension.
In one example of the present disclosure, an interpretation model may be used to determine the prediction result contribution of various decision feature variables in the predictive model under the user feature data. The interpretation model comprises one of the following interpretation models: the Shap value model, the LIME model, and the DeepLift model. How to use the interpretation model to determine the prediction result contribution of each decision characteristic variable in the prediction model under the user characteristic data will be described below with reference to fig. 2.
After the prediction result contribution of each decision characteristic variable is determined as described above, at block 120, a decision characteristic variable combination to be optimized is constructed based on the determined prediction result contribution of each decision characteristic variable. How to construct the decision feature variable combination to be optimized based on the determined prediction result contribution degrees of the respective decision feature variables will be explained below with reference to an example shown in fig. 3.
After the decision feature variable combination to be optimized is constructed, in block 130, optimization processing is performed on the variable values of each decision feature variable in the constructed decision feature variable combination, so that the corresponding prediction result of the prediction model is optimal.
For example, in one example, the optimizing the variable values of the decision characteristic variables in the constructed decision characteristic variable combination may include: optimizing the variable values of each decision characteristic variable in the constructed decision characteristic variable combination by using one of the following optimization algorithms: particle swarm algorithm, genetic algorithm and annealing algorithm. Here, the optimization algorithm used may be determined based on the type of decision feature variable to be subjected to the optimization process. For example, in the case that the decision characteristic variable to be optimized is a classification variable, the optimization algorithm preferably uses a particle swarm algorithm, such as a discrete particle swarm algorithm.
In addition, in another example of the present disclosure, the optimizing the variable values of the decision characteristic variables in the constructed decision characteristic variable combination may include: and in a preset decision variable value range, optimizing the variable value of each decision characteristic variable in the constructed decision characteristic variable combination. In addition, in order to make the optimization effect better, the value range of the decision variable can be properly adjusted according to the actual situation.
After determining the variable values of each decision characteristic variable in the decision characteristic variable combination, at block 140, a decision optimization process is performed according to the variable values of each decision characteristic variable in the decision characteristic variable combination obtained after the optimization process. For example, the decision mechanism used in the decision engine is optimally adjusted by using the variable values of the decision characteristic variables.
Fig. 2 shows a flowchart of one example of a prediction result contribution degree determination process of a decision feature variable according to an embodiment of the present disclosure.
As shown in fig. 2, for each decision feature variable in the predictive model, first, at block 210, one decision feature variable is selected as the initial current decision feature variable. For example, one decision feature variable may be randomly selected from the various decision feature variables as the initial current decision feature variable.
Then, for the current decision feature variable, the operations of blocks 220 to 270 are performed until the corresponding prediction result contribution is determined for all decision feature variables. Specifically, at block 220, the prediction results corresponding to any combination of the remaining decision characteristic variables in the decision characteristic variable combination other than the current decision characteristic variable are determined. For example, assuming that there are 4 decision feature variables V1, V2, V3, V4, and V1 as current decision feature variables in the prediction model, the prediction model is used to calculate a model prediction result in the case where the input of the prediction model is user feature data corresponding to any feature combination of V2, V3, and V4. Next, at block 230, the prediction results corresponding to any feature variable combination of the decision feature variable combinations that includes the current decision feature variable are determined. The feature variable combination described here is a feature variable combination obtained by adding the feature variable V1 for each feature combination in the block 220. Then, at block 240, the difference between the model prediction result determined at block 220 (without the current decision characteristic variable) and the corresponding model prediction result determined at block 230 (with the current decision characteristic variable) is calculated.
After the differences between the model predictions are determined, at block 350, the differences between all the model predictions are averaged to obtain the prediction contribution for the current decision feature variable. Then, at block 260, it is determined whether there are unprocessed decision feature variables. If so, then in block 270, a decision feature variable is selected from the unprocessed decision feature variables as the next current decision feature variable, and then back to block 220, the operations of blocks 220 through 270 are re-performed.
FIG. 3 shows a flow diagram of one example of a decision feature variable combination construction process, according to an embodiment of the present disclosure.
As shown in FIG. 3, first, at block 310, the decision feature variables are ranked based on their determined prediction result contribution. Then, at block 320, a predetermined number of decision characteristic variables with a greater contribution are selected from the sorted decision characteristic variables as the decision characteristic variable combinations to be optimized. Here, the predetermined number may be determined based on an actual application scenario or other suitable conditions. Furthermore, in other examples of the present disclosure, other suitable ways may also be employed to construct the decision feature variable combinations to be optimized.
With the decision optimization method of the present disclosure, for different users, the prediction result contribution of each decision characteristic variable in the prediction model is determined by using the user characteristic data of the user, a decision characteristic variable combination to be optimized is constructed based on the determined prediction results, and a variable value of each decision characteristic variable in the decision characteristic variable combination is determined using an optimization algorithm. In the decision optimization method, different prediction result contribution degrees of each decision characteristic variable can be obtained according to different user characteristic data, so that different decision characteristic variable combinations can be constructed, the optimized decision characteristic variable combinations are different according to different users, and further personalized decision optimization is realized.
By using the decision optimization method disclosed by the invention, the value range of the decision characteristic variable is properly adjusted according to the actual situation, and in the value range of the decision characteristic variable, the optimization processing is carried out on the variable value of each decision characteristic variable in the constructed decision characteristic variable combination, so that the optimization effect can be improved, and the decision optimization effect is further improved.
Fig. 4 shows a flow diagram of one example of an online switching process of a personalized decision mechanism, according to an embodiment of the disclosure.
As shown in fig. 4, after the optimized personalized decision making mechanism and engine are determined as above, when an online decision is made, the decision task is randomly allocated to the original decision making mechanism or the optimized personalized decision making mechanism for processing the incoming customer data to be decided, so as to make a decision, and record the corresponding decision making effect. After the operation is carried out for a period of time, the decision effect of the original decision mechanism or the optimized personalized decision mechanism is evaluated, if the decision effect of the original decision mechanism is better, the original decision mechanism is kept, and if the optimized personalized decision mechanism is better, the optimized personalized decision mechanism is used for replacing the original decision mechanism, so that the stability of decision service is ensured.
Fig. 5 shows a block diagram of a decision optimization apparatus 500 according to an embodiment of the present disclosure. As shown in fig. 5, the decision optimization apparatus 500 includes a contribution degree determination unit 510, a decision characteristic variable combination construction unit 520, an optimization processing unit 530, and a decision optimization unit 540.
The contribution degree determination unit 510 is configured to determine a prediction result contribution degree of each decision characteristic variable in a prediction model under the user characteristic data, the prediction model being created based on the decision target, the prediction model including the decision characteristic variable and the non-decision characteristic variable. The operation of the contribution degree determining unit 510 may refer to the operation of the block 110 described above with reference to fig. 1 and the operation described with reference to fig. 2.
The decision feature variable combination construction unit 520 is configured to construct a decision feature variable combination to be optimized based on the determined prediction result contribution degrees of the respective decision feature variables. The operation of the decision feature variable combination construction unit 520 may refer to the operation of the block 120 described above with reference to fig. 1 and the operation described with reference to fig. 3.
The optimization processing unit 530 is configured to perform optimization processing on the variable values of each decision characteristic variable in the constructed decision characteristic variable combination so as to optimize the corresponding prediction result of the prediction model. In one example of the present disclosure, the optimization processing unit 530 is configured to: and optimizing the variable values of the decision characteristic variables in the constructed decision characteristic variable combination by using one of the following optimization algorithms: particle swarm algorithm, genetic algorithm and annealing algorithm. In another example of the present disclosure, the optimization processing unit 530 is configured to: and in a preset decision variable value range, optimizing the variable value of each decision characteristic variable in the constructed decision characteristic variable combination. The operation of the optimization processing unit 530 may refer to the operation of block 130 described above with reference to fig. 1.
The decision optimization unit 540 is configured to perform decision optimization processing according to the variable values of the decision characteristic variables in the decision characteristic variable combination obtained after the optimization processing. The operation of the decision optimization unit 540 may refer to the operation of block 140 described above with reference to fig. 1.
In one example of the present disclosure, the contribution degree determining unit 510 may be configured to: and determining the prediction result contribution degree of each decision characteristic variable in the prediction model under the user characteristic data by using the interpretation model. The interpretation model may comprise one of the following interpretation models: the Shap value model, the LIME model, and the DeepLift model.
Fig. 6 shows a block diagram of one example of a decision feature variable combination construction unit 520 according to an embodiment of the present disclosure. As shown in fig. 6, the decision feature variable combination construction unit 520 includes a sorting module 521 and a feature selection module 523.
The ranking module 521 is configured to rank the respective decision characteristic variables based on the determined prediction result contribution degrees of the respective decision characteristic variables.
The feature selection module 523 is configured to select a predetermined number of decision feature variables with a larger contribution from the sorted decision feature variables as the decision feature variable combination to be optimized.
Embodiments of a decision optimization method and apparatus according to the present disclosure are described above with reference to fig. 1 to 6. The decision optimization device can be implemented by hardware, software, or a combination of hardware and software.
Fig. 7 illustrates a hardware block diagram of a computing device 700 for decision optimization, according to an embodiment of the disclosure. As shown in fig. 7, computing device 700 may include at least one processor 710, storage 720, memory 730, and communication interface 740, and at least one processor 710, storage 720, memory 730, and communication interface 740 are connected together via a bus 760. The at least one processor 710 executes at least one computer-readable instruction (i.e., the elements described above as being implemented in software) stored or encoded in memory.
In one embodiment, computer-executable instructions are stored in the memory that, when executed, cause the at least one processor 710 to: determining the contribution degree of a prediction result of each decision characteristic variable in a prediction model under user characteristic data, wherein the prediction model is established based on a decision target and comprises the decision characteristic variable and a non-decision characteristic variable; constructing a decision characteristic variable combination to be optimized based on the determined prediction result contribution of each decision characteristic variable; optimizing the variable values of each decision characteristic variable in the constructed decision characteristic variable combination so as to ensure that the corresponding prediction result of the prediction model is optimal; and performing decision optimization processing according to the variable values of the decision characteristic variables in the decision characteristic variable combination obtained after optimization processing.
It should be understood that the computer-executable instructions stored in the memory, when executed, cause the at least one processor 710 to perform the various operations and functions described above in connection with fig. 1-6 in the various embodiments of the present disclosure.
In the present disclosure, computing device 700 may include, but is not limited to: personal computers, server computers, workstations, desktop computers, laptop computers, notebook computers, mobile computing devices, smart phones, tablet computers, cellular phones, Personal Digital Assistants (PDAs), handheld devices, messaging devices, wearable computing devices, consumer electronics, and so forth.
According to one embodiment, a program product, such as a non-transitory machine-readable medium, is provided. A non-transitory machine-readable medium may have instructions (i.e., elements described above as being implemented in software) that, when executed by a machine, cause the machine to perform various operations and functions described above in connection with fig. 1-6 in various embodiments of the present disclosure. Specifically, a system or apparatus may be provided which is provided with a readable storage medium on which software program code implementing the functions of any of the above embodiments is stored, and causes a computer or processor of the system or apparatus to read out and execute instructions stored in the readable storage medium.
According to one embodiment, a program product, such as a non-transitory machine-readable medium, is provided. A non-transitory machine-readable medium may have instructions (i.e., elements described above as being implemented in software) that, when executed by a machine, cause the machine to perform various operations and functions described above in connection with fig. 1-6 in various embodiments of the present disclosure. Specifically, a system or apparatus may be provided which is provided with a readable storage medium on which software program code implementing the functions of any of the above embodiments is stored, and causes a computer or processor of the system or apparatus to read out and execute instructions stored in the readable storage medium.
In this case, the program code itself read from the readable medium can realize the functions of any of the above-described embodiments, and thus the machine-readable code and the readable storage medium storing the machine-readable code form part of the present invention.
Examples of the readable storage medium include floppy disks, hard disks, magneto-optical disks, optical disks (e.g., CD-ROMs, CD-R, CD-RWs, DVD-ROMs, DVD-RAMs, DVD-RWs), magnetic tapes, nonvolatile memory cards, and ROMs. Alternatively, the program code may be downloaded from a server computer or from the cloud via a communications network.
It will be understood by those skilled in the art that various changes and modifications may be made in the above-disclosed embodiments without departing from the spirit of the invention. Accordingly, the scope of the invention should be determined from the following claims.
It should be noted that not all steps and units in the above flows and system structure diagrams are necessary, and some steps or units may be omitted according to actual needs. The execution order of the steps is not fixed, and can be determined as required. The apparatus structures described in the above embodiments may be physical structures or logical structures, that is, some units may be implemented by the same physical entity, or some units may be implemented by a plurality of physical entities, or some units may be implemented by some components in a plurality of independent devices.
In the above embodiments, the hardware units or modules may be implemented mechanically or electrically. For example, a hardware unit, module or processor may comprise permanently dedicated circuitry or logic (such as a dedicated processor, FPGA or ASIC) to perform the corresponding operations. The hardware units or processors may also include programmable logic or circuitry (e.g., a general purpose processor or other programmable processor) that may be temporarily configured by software to perform the corresponding operations. The specific implementation (mechanical, or dedicated permanent, or temporarily set) may be determined based on cost and time considerations.
The detailed description set forth above in connection with the appended drawings describes exemplary embodiments but does not represent all embodiments that may be practiced or fall within the scope of the claims. The term "exemplary" used throughout this specification means "serving as an example, instance, or illustration," and does not mean "preferred" or "advantageous" over other embodiments. The detailed description includes specific details for the purpose of providing an understanding of the described technology. However, the techniques may be practiced without these specific details. In some instances, well-known structures and devices are shown in block diagram form in order to avoid obscuring the concepts of the described embodiments.
The previous description of the disclosure is provided to enable any person skilled in the art to make or use the disclosure. Various modifications to the disclosure will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other variations without departing from the scope of the disclosure. Thus, the disclosure is not intended to be limited to the examples and designs described herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (14)
1. A method for decision optimization, comprising:
determining the contribution of each decision characteristic variable in a prediction model under user characteristic data, wherein the prediction model is created based on a decision target and comprises decision characteristic variables and non-decision characteristic variables, the decision characteristic variables refer to characteristic variables which can be interfered and changed by a decision party, the non-decision characteristic variables refer to characteristic variables which cannot be interfered by the decision party, and the user characteristic data are user characteristic data corresponding to the decision characteristic variables and the non-decision characteristic variables;
selecting a decision characteristic variable combination to be optimized from each decision characteristic variable based on the determined prediction result contribution degree of each decision characteristic variable;
optimizing the variable values of the decision characteristic variables in the selected decision characteristic variable combination so as to optimize the corresponding prediction result of the prediction model; and
and performing decision optimization processing according to the variable values of the decision characteristic variables in the decision characteristic variable combination obtained after optimization processing.
2. The method of claim 1, wherein determining a prediction result contribution of each decision feature variable in the predictive model under the user feature data comprises:
and determining the prediction result contribution degree of each decision characteristic variable in the prediction model under the user characteristic data by using the interpretation model.
3. The method of claim 2, wherein the interpretation model comprises one of the following interpretation models:
the Shap value model, the LIME model, and the DeepLift model.
4. The method of claim 1, wherein selecting a decision feature variable combination to be optimized from the respective decision feature variables based on the determined prediction result contribution of the respective decision feature variables comprises:
based on the determined contribution degree of the prediction result of each decision characteristic variable, sequencing each decision characteristic variable; and
and selecting a preset number of decision characteristic variables with higher contribution degree from the sorted decision characteristic variables as a decision characteristic variable combination to be optimized.
5. The method of claim 1, wherein optimizing the variable values of each decision feature variable in the selected decision feature variable combination comprises:
optimizing the variable values of the decision characteristic variables in the selected decision characteristic variable combination by using one of the following optimization algorithms:
particle swarm algorithm, genetic algorithm and annealing algorithm.
6. The method of claim 1, wherein optimizing the variable values of each decision feature variable in the selected decision feature variable combination comprises:
and in a preset decision variable value range, optimizing the variable values of the decision characteristic variables in the selected decision characteristic variable combination.
7. An apparatus for decision optimization, comprising:
the contribution degree determining unit is configured to determine the contribution degree of a prediction result of each decision characteristic variable in a prediction model under user characteristic data, the prediction model is created based on a decision target, the prediction model comprises decision characteristic variables and non-decision characteristic variables, the decision characteristic variables refer to characteristic variables which can be changed by a decision party in an interference mode, the non-decision characteristic variables refer to characteristic variables which cannot be changed by the decision party in the interference mode, and the user characteristic data correspond to the decision characteristic variables and the non-decision characteristic variables;
a decision characteristic variable combination construction unit configured to select a decision characteristic variable combination to be optimized from each decision characteristic variable based on the determined prediction result contribution degree of each decision characteristic variable;
an optimization processing unit configured to perform optimization processing on the variable values of each decision characteristic variable in the selected decision characteristic variable combination so as to optimize the corresponding prediction result of the prediction model; and
and the decision optimization unit is configured to perform decision optimization processing according to the variable values of the decision characteristic variables in the decision characteristic variable combination obtained after optimization processing.
8. The apparatus of claim 7, wherein the contribution degree determination unit is configured to:
and determining the prediction result contribution degree of each decision characteristic variable in the prediction model under the user characteristic data by using the interpretation model.
9. The apparatus of claim 8, wherein the interpretation model comprises one of:
the Shap value model, the LIME model, and the DeepLift model.
10. The apparatus of claim 7, wherein the decision feature variable combination construction unit comprises:
the sequencing module is configured to sequence the decision characteristic variables based on the determined prediction result contribution degrees of the decision characteristic variables; and
and the characteristic selection module is configured to select a predetermined number of decision characteristic variables with higher contribution degree from the sorted decision characteristic variables as the decision characteristic variable combination to be optimized.
11. The apparatus of claim 7, wherein the optimization processing unit is configured to:
optimizing the variable values of the decision characteristic variables in the selected decision characteristic variable combination by using one of the following optimization algorithms:
particle swarm algorithm, genetic algorithm and annealing algorithm.
12. The apparatus of claim 7, wherein the optimization processing unit is configured to:
and in a preset decision variable value range, optimizing the variable values of the decision characteristic variables in the selected decision characteristic variable combination.
13. A computing device, comprising:
at least one processor, and
a memory coupled with the at least one processor, the memory storing instructions that, when executed by the at least one processor, cause the at least one processor to perform the method of any of claims 1-6.
14. A non-transitory machine-readable storage medium storing executable instructions that, when executed, cause the machine to perform the method of any of claims 1-6.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910176389.8A CN110046799B (en) | 2019-03-08 | 2019-03-08 | Decision optimization method and device |
| PCT/CN2020/070492 WO2020181907A1 (en) | 2019-03-08 | 2020-01-06 | Decision-making optimization method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910176389.8A CN110046799B (en) | 2019-03-08 | 2019-03-08 | Decision optimization method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110046799A CN110046799A (en) | 2019-07-23 |
| CN110046799B true CN110046799B (en) | 2021-09-10 |
Family
ID=67274727
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910176389.8A Active CN110046799B (en) | 2019-03-08 | 2019-03-08 | Decision optimization method and device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN110046799B (en) |
| WO (1) | WO2020181907A1 (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110046799B (en) * | 2019-03-08 | 2021-09-10 | 创新先进技术有限公司 | Decision optimization method and device |
| CN111461892B (en) * | 2020-03-31 | 2021-07-06 | 支付宝(杭州)信息技术有限公司 | Method and device for selecting derived variables of risk identification model |
| CN111262887B (en) * | 2020-04-26 | 2020-08-28 | 腾讯科技（深圳）有限公司 | Network risk detection method, device, equipment and medium based on object characteristics |
| CN112766586B (en) * | 2021-01-26 | 2024-07-09 | 绍兴文理学院 | Multi-target optimization examination method and system based on step-by-step progressive strategy |
| CN113570260A (en) * | 2021-07-30 | 2021-10-29 | 北京房江湖科技有限公司 | Task allocation method, computer-readable storage medium and electronic device |
| CN113590691B (en) * | 2021-08-04 | 2024-06-28 | 浙江网商银行股份有限公司 | Target object processing method and device |
| CN113902450A (en) * | 2021-12-08 | 2022-01-07 | 四川新网银行股份有限公司 | Interpretable fraud transaction identification method and device |
| CN117710066B (en) * | 2024-02-05 | 2024-05-10 | 厦门傲凡科技股份有限公司 | Financial customer recommendation method and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106022473A (en) * | 2016-05-23 | 2016-10-12 | 大连理工大学 | Construction method for gene regulatory network by combining particle swarm optimization (PSO) with genetic algorithm |
| CN107703749A (en) * | 2017-10-10 | 2018-02-16 | 常州大学 | A kind of pid parameter optimization method based on self-adapted genetic algorithm |
| CN108665277A (en) * | 2017-03-27 | 2018-10-16 | 阿里巴巴集团控股有限公司 | A kind of information processing method and device |
| CN109243172A (en) * | 2018-07-25 | 2019-01-18 | 华南理工大学 | Traffic flow forecasting method based on genetic algorithm optimization LSTM neural network |
| CN109271695A (en) * | 2018-09-07 | 2019-01-25 | 中南大学 | Multiple target antenna design method neural network based |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109036568A (en) * | 2018-09-03 | 2018-12-18 | 浪潮软件集团有限公司 | Method for establishing prediction model based on naive Bayes algorithm |
| CN109343019B (en) * | 2018-11-05 | 2019-10-18 | 中国矿业大学(北京) | A kind of Georadar Data means of interpretation and device |
| CN110046799B (en) * | 2019-03-08 | 2021-09-10 | 创新先进技术有限公司 | Decision optimization method and device |
-
2019
- 2019-03-08 CN CN201910176389.8A patent/CN110046799B/en active Active
-
2020
- 2020-01-06 WO PCT/CN2020/070492 patent/WO2020181907A1/en active Application Filing
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106022473A (en) * | 2016-05-23 | 2016-10-12 | 大连理工大学 | Construction method for gene regulatory network by combining particle swarm optimization (PSO) with genetic algorithm |
| CN108665277A (en) * | 2017-03-27 | 2018-10-16 | 阿里巴巴集团控股有限公司 | A kind of information processing method and device |
| CN107703749A (en) * | 2017-10-10 | 2018-02-16 | 常州大学 | A kind of pid parameter optimization method based on self-adapted genetic algorithm |
| CN109243172A (en) * | 2018-07-25 | 2019-01-18 | 华南理工大学 | Traffic flow forecasting method based on genetic algorithm optimization LSTM neural network |
| CN109271695A (en) * | 2018-09-07 | 2019-01-25 | 中南大学 | Multiple target antenna design method neural network based |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110046799A (en) | 2019-07-23 |
| WO2020181907A1 (en) | 2020-09-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110046799B (en) | Decision optimization method and device | |
| JP6912591B2 (en) | Risk management control methods and devices | |
| CN106355391B (en) | Service processing method and device | |
| CN111027715B (en) | Monte Carlo-based federated learning model training method and device | |
| US20190035015A1 (en) | Method and apparatus for obtaining a stable credit score | |
| CN115526363A (en) | Business data processing method and device, computer equipment and storage medium | |
| CN104965844A (en) | Information processing method and apparatus | |
| JP6253744B1 (en) | Information analysis apparatus, information analysis method, and information analysis program | |
| JP6253745B1 (en) | Information analysis apparatus, information analysis method, and information analysis program | |
| US20220405621A1 (en) | Controlling reachability in a collaboratively filtered recommender | |
| CN107038165B (en) | Service parameter acquisition method and device | |
| CN113902473A (en) | Training method and device of business prediction system | |
| CN109714301A (en) | Register Risk Identification Method, device, electronic equipment and storage medium | |
| CN111324370A (en) | Method and device for carrying out risk processing on to-be-on-line small program | |
| CN112669143A (en) | Risk assessment method, device and equipment based on associated network and storage medium | |
| Valizadegan et al. | Learning to trade off between exploration and exploitation in multiclass bandit prediction | |
| CN112926090A (en) | Service analysis method and device based on differential privacy | |
| CN113033824A (en) | Model hyper-parameter determination method, model training method and system | |
| CN112967044B (en) | Payment service processing method and device | |
| Pelegrina et al. | Analysis of trade-offs in fair principal component analysis based on multi-objective optimization | |
| CN109685091A (en) | It is determined using the number experience target of Bayes's mode | |
| CN114925275A (en) | Product recommendation method and device, computer equipment and storage medium | |
| CN110020728B (en) | Service model reinforcement learning method and device | |
| Guo et al. | Fair learning to rank with distribution-free risk control | |
| Higuchi et al. | Learning Context-dependent Personal Preferences for Adaptive Recommendation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20201019 Address after: English genus Applicant after: Innovative advanced technology Co.,Ltd. Address before: English genus Applicant before: Advanced innovation technology Co.,Ltd. Effective date of registration: 20201019 Address after: English genus Applicant after: Advanced innovation technology Co.,Ltd. Address before: A four-storey 847 mailbox in Grand Cayman Capital Building, British Cayman Islands Applicant before: Alibaba Group Holding Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |