Automatic strategy mining method and device based on multi-dimensional data driving, equipment and medium

Abstract

The invention discloses an automatic strategy mining method and device based on multi-dimensional data driving, equipment and a medium. The method comprises the steps that a feature data set, a label data set, an evaluation index data set and an actual decision result data set in a target time period in the financial field are acquired; dividing the spliced data set into a training set, a test set and a verification set; determining a target function; determining a hit rate and a recall rate of each branch in each decision tree; screening the corresponding branches to obtain a plurality of optimal branches, and removing data of the optimal branches in the training set to obtain a target strategy set; evaluating and screening the target strategy set; determining weights of a plurality of preset evaluation indexes according to the spliced data set; and determining the score of each strategy in the target strategy set according to a plurality of preset evaluation indexes and the weight corresponding to each preset evaluation index, and obtaining an optimal strategy. The invention belongs to the field of financial strategy mining. The automation degree of strategy mining can be improved.

Description

Technical Field

[0001] The present invention relates to the field of financial strategy mining, and in particular to a multi-dimensional data driven automatic strategy mining method, device, equipment and medium. Background Art

[0002] Financial strategy mining refers to the process of identifying and extracting patterns, rules or strategies that can help investors or financial institutions make better decisions from a large amount of financial data and information through data analysis, machine learning, natural language processing and other technical means. Strategies can be used to guide investment behavior, optimize risk management, improve business processes and enhance customer service.

[0003] Traditional mining methods usually require manual parameter setting and feature selection. The process not only consumes a lot of manpower and time resources, but also is likely to introduce deviations caused by human factors due to differences in subjective cognition. When faced with complex multidimensional data, the mining efficiency and accuracy of traditional methods are often greatly limited. Therefore, a new strategy automatic mining method is urgently needed. Summary of the invention

[0004] The present invention solves the technical problem of low efficiency of strategy mining in the prior art by providing a multi-dimensional data driven automatic strategy mining method, device, equipment and medium, and achieves the technical effect of improving the efficiency of strategy mining.

[0005] In a first aspect, the present invention provides a multi-dimensional data-driven automatic strategy mining method, the method comprising:

[0006] Obtain feature data sets, label data sets, evaluation index data sets, and actual decision-making result data sets in the financial field within the target time period;

[0007] The feature data set, the label data set, the evaluation index data set and the actual decision result data set are spliced ​​to obtain a spliced ​​data set, and the spliced ​​data set is divided into a training set, a test set and a validation set;

[0008] Determine the objective function, which is used to determine the optimization direction and optimization goal;

[0009] Build several decision trees based on the training set and determine the hit rate and recall rate of each branch in each decision tree, where one branch corresponds to one strategy;

[0010] Based on the preset hit rate threshold, the hit rate and recall rate corresponding to each branch, the corresponding branches are screened to obtain several optimal branches, and the data of the optimal branches in the training set are eliminated to obtain the target strategy set;

[0011] Evaluate and screen the target strategy set based on the test set and validation set;

[0012] Determine the weights of several preset evaluation indicators based on the spliced ​​data set;

[0013] According to several preset evaluation indicators and the weights corresponding to each preset evaluation indicator, the score of each strategy in the target strategy set is determined, and the optimal strategy is obtained.

[0014] Further, based on the preset hit rate threshold, the hit rate and recall rate corresponding to each branch, the corresponding branches are screened to obtain several optimal branches, and the data of the optimal branches in the training set are removed to obtain the target strategy set, including steps S151-S152:

[0015] Step S151, based on a preset hit rate threshold, the hit rate and recall rate corresponding to each branch, the corresponding branch is screened to determine whether the branch is the optimal branch;

[0016] Step S152: if the branch is the optimal branch, remove the data of the branch in the training set and include the branch in the target strategy set; if it is not the optimal branch, re-execute step S151 with other branches.

[0017] Furthermore, based on a preset hit rate threshold, the hit rate and recall rate corresponding to each branch, the corresponding branches are screened, including:

[0018]

[0019] Where SelectBranches(T,θ,N) is the screening function, T is the decision tree, b and b′ are both branches in the decision tree T, HitRate(b) is the hit rate of branch (b), Recall(b) is the recall rate of branch (b), Recall(b′) is the recall rate of branch (b′), θ is the preset hit rate threshold, and N is the number of branches.

[0020] Furthermore, the objective function includes:

[0021] ObjectiveFunction=maxR(Parameters)

[0022] HitRate<θ

[0023] Among them, ObjectiveFunction is the objective function, R(Parameters) is the recall rate as a function of the strategy, Parameters is the strategy set, HitRate is the hit rate, and θ is the preset hit rate threshold.

[0024] Furthermore, the concatenated dataset is divided into a training set, a test set, and a validation set, including:

[0025] Dividing the target time period into a first time period and a second time period, wherein the time period length of the first time period is greater than the time period length of the second time period and the first time period is before the second time period;

[0026] Dividing the spliced ​​data in the first time period into a training set and a test set according to a preset ratio;

[0027] The concatenated data in the second time period is divided into a validation set.

[0028] Furthermore, the optimal strategy is obtained, including:

[0029] The strategy corresponding to the highest score is regarded as the optimal strategy.

[0030] Furthermore, the score of each strategy in the target strategy set is determined, including:

[0031]

[0032] Among them, F(s k ) is strategy s k The score, w i is the weight of the i-th evaluation index, x ki For strategies k The i-th evaluation index of .

[0033] In a second aspect, the present invention provides a strategy automatic mining device based on multidimensional data drive, the device comprising:

[0034] The acquisition module is used to obtain the feature data set, label data set, evaluation index data set and actual decision result data set in the financial field within the target time period;

[0035] A partitioning module is used to concatenate the feature data set, the label data set, the evaluation index data set, and the actual decision result data set to obtain a concatenated data set, and to divide the concatenated data set into a training set, a test set, and a validation set;

[0036] Function determination module, used to determine the objective function, which is used to determine the optimization direction and optimization target;

[0037] The decision module is used to construct several decision trees based on the training set and determine the hit rate and recall rate of each branch in each decision tree, where one branch corresponds to one strategy;

[0038] The optimization module is used to screen the corresponding branches based on the preset hit rate threshold, the hit rate and recall rate corresponding to each branch, obtain several optimal branches, and eliminate the data of the optimal branches in the training set to obtain the target strategy set;

[0039] The screening module is used to evaluate and screen the target strategy set based on the test set and the validation set;

[0040] A weight module, used to determine the weights of several preset evaluation indicators according to the spliced ​​data set;

[0041] The optimal strategy module is used to determine the score of each strategy in the target strategy set according to a number of preset evaluation indicators and the weights corresponding to each preset evaluation indicator, and obtain the optimal strategy.

[0042] In a third aspect, the present invention provides an electronic device, comprising:

[0043] processor;

[0044] a memory for storing processor-executable instructions;

[0045] The processor is configured to execute to implement the multi-dimensional data driven automatic strategy mining method provided in the first aspect.

[0046] In a fourth aspect, the present invention provides a non-temporary computer-readable storage medium, which, when the instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to implement the multidimensional data-driven automatic strategy mining method provided in the first aspect.

[0047] One or more technical solutions provided in the present invention have at least the following technical effects or advantages:

[0048] The present invention provides a multi-dimensional data-driven automatic strategy mining method, which comprises: obtaining a feature data set, a label data set, an evaluation index data set and an actual decision result data set in a financial field within a target time period; splicing the feature data set, the label data set, the evaluation index data set and the actual decision result data set to obtain a spliced ​​data set, and dividing the spliced ​​data set into a training set, a test set and a verification set; determining an objective function, which is used to determine an optimization direction and an optimization target; constructing a plurality of decision trees according to the training set, and determining a hit rate and a recall rate of each branch in each decision tree, wherein one branch corresponds to one strategy; based on a preset hit rate threshold, the hit rate and the recall rate corresponding to each branch, screening the corresponding branches to obtain a plurality of optimal branches, and eliminating the data of the optimal branch in the training set to obtain a target strategy set; evaluating and screening the target strategy set based on a test set and a verification set; determining the weights of a plurality of preset evaluation indicators according to the spliced ​​data set; determining the score of each strategy in the target strategy set according to the plurality of preset evaluation indicators and the weights corresponding to each preset evaluation indicator, and obtaining the optimal strategy. The present invention constructs an automated strategy mining mechanism that can automatically adjust the strategy set according to multidimensional data, achieve the global optimal recall rate under the requirement of setting the hit rate, without manual intervention, and improve the automation level of strategy mining. Pruning operations are performed on the strategy set based on business experience to remove parts that do not meet actual business needs. Multiple evaluation dimensions are constructed based on evaluation indicators, and the strategy set that performs best on multiple evaluation indicators is screened out from a large number of strategy sets. By predicting, measuring and comparing the actual impact of decisions of different strategy sets on production, the most powerful strategy for the actual goal is selected. The present invention is based on a decision tree and has interpretability, which improves the efficiency and accuracy of strategy updates and enables strategies to adapt to rapidly changing data environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0050] Figure 1 A schematic flow chart of the automatic strategy mining method based on multidimensional data drive provided by the present invention. DETAILED DESCRIPTION

[0051] The embodiment of the present invention solves the technical problem of low efficiency of strategy mining in the prior art by providing a strategy automatic mining method based on multi-dimensional data drive.

[0052] The technical solution of the present invention is to solve the above technical problems, and the overall idea is as follows:

[0053] The invention discloses an automatic strategy mining method based on multidimensional data drive, the method comprising: obtaining a feature data set, a label data set, an evaluation index data set and an actual decision result data set in the financial field within a target time period; splicing the feature data set, the label data set, the evaluation index data set and the actual decision result data set to obtain a spliced ​​data set, and dividing the spliced ​​data set into a training set, a test set and a validation set; determining an objective function, the objective function is used to determine an optimization direction and an optimization target; constructing a plurality of decision trees according to the training set, and determining a hit rate and a recall rate of each branch in each decision tree, wherein one branch corresponds to one strategy; based on a preset hit rate threshold, the hit rate and the recall rate corresponding to each branch, screening the corresponding branches to obtain a plurality of optimal branches, and eliminating the data of the optimal branch in the training set to obtain a target strategy set; evaluating and screening the target strategy set based on the test set and the validation set; determining the weights of a plurality of preset evaluation indicators according to the spliced ​​data set; determining the score of each strategy in the target strategy set according to a plurality of preset evaluation indicators and the weights corresponding to each preset evaluation indicator, and obtaining the optimal strategy.

[0054] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0055] First of all, the term "and / or" in this article is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0056] The present invention provides Figure 1 The strategy automatic mining method based on multidimensional data driving shown includes steps S11-S18:

[0057] Step S11, obtaining a feature data set, a label data set, an evaluation index data set, and an actual decision result data set in the financial field within a target time period.

[0058] The target time period can be selected according to actual conditions. The present invention takes the data in the financial field from January 1, 2022 to August 1, 2024 as an example for explanation.

[0059] A feature dataset is a collection of data containing multiple types of features. Feature datasets are used to describe the attributes and characteristics of samples. In financial scenarios, feature datasets may include but are not limited to GPS location feature data, personal information feature data, aggregate feature data, and image feature data. By constructing feature datasets, the attributes and characteristics of samples can be fully described, providing a reliable data foundation for fraud detection and risk assessment in financial scenarios.

[0060] The label dataset is used to define the label set to identify whether the samples in the dataset hit a certain fraud or overdue scenario. For example, in the anti-fraud model, the label dataset can contain a binary label to indicate whether the sample was defrauded offline. In addition, the label dataset can also use a customized combination of multiple labels to more accurately describe the fraud risk of the sample.

[0061] The evaluation index dataset is a dataset containing multiple types of indicators, which is used to describe the risk and fraud in financial scenarios. In financial scenarios, the evaluation indicators can include overdue rate, fraud overdue rate, and abnormal case investigation rate.

[0062] The actual decision result dataset is used to describe the decision results in financial scenarios. In financial scenarios, the actual decision results may include anti-fraud rejection, manual review and approval, etc.

[0063] It can be understood that within the target time period, a number of data in the financial field are organized into corresponding feature data sets, label data sets, evaluation index data sets and actual decision result data sets.

[0064] Step S12, concatenating the feature data set, the label data set, the evaluation index data set and the actual decision result data set to obtain a concatenated data set, and dividing the concatenated data set into a training set, a test set and a validation set.

[0065] Specifically, the method includes: dividing the target time period into a first time period and a second time period, wherein the time period length of the first time period is greater than the time period length of the second time period and the first time period is before the second time period; dividing the spliced ​​data in the first time period into a training set and a test set according to a preset ratio; and dividing the spliced ​​data in the second time period into a verification set.

[0066] The purpose of dividing the time periods in the present invention is to ensure the stability of the strategy in different time intervals. The data in the second time period is more in line with the current time, so based on the data in the second time period, the verification set is divided to make the verification result more accurate.

[0067] For example, for a spliced ​​data set from January 1, 2022 to November 1, 2024, it can be divided into January 1, 2022 to August 1, 2024 (first time period) and August 1, 2024 to November 1, 2024 (second time period).

[0068] The preset ratio can be determined according to actual conditions, for example, 7:3, in which case 70% of the data in the first time period is used as a training set and 30% of the data is used as a test set.

[0069] Step S13, determining the objective function, which is used to determine the optimization direction and optimization target.

[0070] Use the training set for strategy mining, that is, to discover and extract effective strategies from the training set by continuously iterating the improved decision tree. Decision tree mining is a data classification and regression method based on a tree structure. Its core principle is to split the data set by recursively selecting the best splitting attribute until a leaf node is generated after the stop condition is met. Each branch of the decision tree can be used as a decision rule or strategy in financial risk assessment, such as deciding whether to approve a loan based on the borrower's credit score, income level, and debt ratio. Each node represents a decision point, and each branch represents a different action path taken based on the attribute value of the decision point.

[0071] Regarding the objective function, including:

[0072] ObjectiveFunction=maxR(Parameters)

[0073] HitRate<θ

[0074] Among them, ObjectiveFunction is the objective function, R(Parameters) is the recall rate as a function of the strategy, Parameters is the strategy set, HitRate is the hit rate, and θ is the preset hit rate threshold.

[0075] Regarding recall, including:

[0076]

[0077] Among them, Recall is the recall rate, TP is the number of samples correctly predicted as positive, and FN is the number of samples incorrectly predicted as negative but actually positive.

[0078]

[0079] Among them, Precision is the hit rate, TP is the number of samples correctly predicted as positive by the model, and FP is the number of samples that are incorrectly predicted as negative and are actually negative.

[0080] Based on the objective function, we can ensure that the strategy can improve the recall rate as much as possible while meeting the hit rate requirements, thereby achieving the optimization efficiency and accuracy of the strategy.

[0081] Step S14, constructing several decision trees according to the training set, and determining the hit rate and recall rate of each branch in each decision tree, wherein one branch corresponds to one strategy.

[0082] Random row sampling and column sampling are performed based on the training set to build multiple decision trees. For each decision tree, the hit rate and recall rate of each branch set can be determined.

[0083] Step S15, based on the preset hit rate threshold, the hit rate and recall rate corresponding to each branch, the corresponding branches are screened to obtain several optimal branches, and the data of the optimal branches in the training set are eliminated to obtain the target strategy set.

[0084] Specifically, it includes steps S151-S152: Step S151, based on the preset hit rate threshold, the hit rate and recall rate corresponding to each branch, the corresponding branches are screened to determine whether the branch is the optimal branch; Step S152, if the branch is the optimal branch, the data of the branch in the training set is removed, and the branch is included in the target strategy set; if it is not the optimal branch, step S151 is re-executed with other branches.

[0085] It should be noted that the data corresponding to the optimal branch (also called data set) is eliminated, and S151-S152 is repeatedly executed until the hit rate is less than the preset hit rate threshold θ and the recall rate no longer increases. After each update, it is still necessary to check whether the hit rate is still less than θ. If so, continue the optimization; if not, stop the optimization.

[0086] The process of selecting the optimal branches can be formalized as the following function, which accepts the decision tree (T), the hit rate threshold (θ), and the number of branches to be selected (N) as input, and returns a list of branches that meet the conditions.

[0087] Based on the preset hit rate threshold, the hit rate and recall rate of each branch, the corresponding branches are screened, including:

[0088]

[0089] Where SelectBranches(T,θ,N) is the screening function, T is the decision tree, b and b′ are both branches in the decision tree T, HitRate(b) is the hit rate of branch (b), Recall(b) is the recall rate of branch (b), Recall(b′) is the recall rate of branch (b′), θ is the preset hit rate threshold, and N is the number of branches.

[0090] The mathematical expression for obtaining the target policy set may include:

[0091]

[0092] Among them, s′ is the target strategy set, S b is the data set that branches b hits, U is the union, and SelectedBranches is the selected optimal branches.

[0093] Step S16, evaluating and screening the target strategy set based on the test set and the validation set.

[0094] Evaluate the performance of the strategy based on the test set. Ensure that the strategy can reach the maximum recall rate while maintaining the hit rate, and calculate the PSI (Population Stability Index) of the time interval. The test set is used to verify whether the strategy can achieve stable results in the same sample set, so as to eliminate unstable strategy sets.

[0095] The validation set is used to screen out the strategies with the best stability and effect. The validation set pays more attention to the stability in the time dimension, thereby eliminating the strategy sets that are unstable in the time dimension. Through the above steps, the automatic mining and updating of strategies are realized, and multiple globally optimal strategy sets can be found, thereby improving the accuracy and efficiency of decision-making.

[0096] In addition, during the policy optimization process, the policy set can be pruned. Specifically, a rule-based approach can be used to set specific rules based on business experience to eliminate policy branches that are too costly to implement or conflict with business common sense.

[0097] Step S17, determining the weights of several preset evaluation indicators according to the spliced ​​data set.

[0098] Based on the evaluation indicators in the validation set (such as overdue rate, fraud overdue rate and case investigation abnormality rate, etc. E = e1, e2, ..., e k ), use the analytic hierarchy process to determine the weight of each indicator. (It can also be used when the performance of the test set and the training set is sufficient)

[0099] Specifically, a hierarchical model is established, with the target layer set as the optimal strategy set, the criterion layer as various evaluation indicators, and the solution layer as different strategy sets. The weights are preset in advance, and the importance of each element at the same level with respect to a criterion at the previous level is compared pairwise to construct a judgment matrix B = (b ij ) n×n , where b ijIndicates the importance of indicator i relative to indicator j. Using the eigenvector method to solve the judgment matrix, we can get the weight of each evaluation indicator w = (w1, w2, ..., w n ),

[0100] Bw=λ max w

[0101] Among them, λ max is the maximum eigenvalue of the judgment matrix. Multiple evaluation dimensions can be constructed based on several weights.

[0102] Step S18, according to a number of preset evaluation indicators and the weights corresponding to the preset evaluation indicators, the score of each strategy in the target strategy set is determined, and the optimal strategy is obtained.

[0103] Obtaining the optimal strategy includes: taking the strategy corresponding to the highest score as the optimal strategy.

[0104] Determine the score for each policy in the target policy set, including:

[0105]

[0106] Among them, F(s k ) is strategy s k The score, w i is the weight of the i-th evaluation index, x ki For strategies k The i-th evaluation index of .

[0107] By establishing a simulation environment and using historical data to run simulations, we can predict the impact of different strategy sets on the decisions made. p ) Calculate the actual hit rate, anti-fraud rejection rate after launch, approval rate and change in transfer rate, measure and compare different strategies, and then carefully select the most effective strategy for the actual goal.

[0108] In summary, the present invention provides a multi-dimensional data-driven automatic strategy mining method, the method comprising: obtaining a feature data set, a label data set, an evaluation index data set and an actual decision result data set in the financial field within a target time period; splicing the feature data set, the label data set, the evaluation index data set and the actual decision result data set to obtain a spliced ​​data set, and dividing the spliced ​​data set into a training set, a test set and a verification set; determining an objective function, the objective function is used to determine an optimization direction and an optimization target; constructing a plurality of decision trees according to the training set, and determining the hit rate and recall rate of each branch in each decision tree, wherein one branch corresponds to one strategy; based on a preset hit rate threshold, the hit rate and recall rate corresponding to each branch, screening the corresponding branches to obtain a plurality of optimal branches, and eliminating the data of the optimal branch in the training set to obtain a target strategy set; evaluating and screening the target strategy set based on the test set and the verification set; determining the weights of a plurality of preset evaluation indicators according to the spliced ​​data set; determining the score of each strategy in the target strategy set according to a plurality of preset evaluation indicators and the weights corresponding to each preset evaluation indicator, and obtaining the optimal strategy. The present invention constructs an automated strategy mining mechanism that can automatically adjust the strategy set according to multidimensional data, achieve the global optimal recall rate under the requirement of setting the hit rate, without manual intervention, and improve the automation level of strategy mining. Pruning operations are performed on the strategy set based on business experience to remove parts that do not meet actual business needs. Multiple evaluation dimensions are constructed based on evaluation indicators, and the strategy set that performs best on multiple evaluation indicators is screened out from a large number of strategy sets. By predicting, measuring and comparing the actual impact of decisions of different strategy sets on production, the most powerful strategy for the actual goal is selected. The present invention is based on a decision tree and has interpretability, which improves the efficiency and accuracy of strategy updates and enables strategies to adapt to rapidly changing data environments.

[0109] Based on the same inventive concept, the present invention provides a strategy automatic mining device driven by multidimensional data, the device comprising:

[0110] The acquisition module is used to obtain the feature data set, label data set, evaluation index data set and actual decision result data set in the financial field within the target time period;

[0111] A partitioning module is used to concatenate the feature data set, the label data set, the evaluation index data set, and the actual decision result data set to obtain a concatenated data set, and to divide the concatenated data set into a training set, a test set, and a validation set;

[0112] Function determination module, used to determine the objective function, which is used to determine the optimization direction and optimization target;

[0113] The decision module is used to construct several decision trees based on the training set and determine the hit rate and recall rate of each branch in each decision tree, where one branch corresponds to one strategy;

[0114] The optimization module is used to screen the corresponding branches based on the preset hit rate threshold, the hit rate and recall rate corresponding to each branch, obtain several optimal branches, and eliminate the data of the optimal branches in the training set to obtain the target strategy set;

[0115] The screening module is used to evaluate and screen the target strategy set based on the test set and the validation set;

[0116] A weight module, used to determine the weights of several preset evaluation indicators according to the spliced ​​data set;

[0117] The optimal strategy module is used to determine the score of each strategy in the target strategy set according to a number of preset evaluation indicators and the weights corresponding to each preset evaluation indicator, and obtain the optimal strategy.

[0118] Based on the same inventive concept, the present invention also provides an electronic device, including:

[0119] processor;

[0120] a memory for storing processor-executable instructions;

[0121] The processor is configured to execute to implement a multi-dimensional data driven strategy automatic mining method as provided above.

[0122] Based on the same inventive concept, the present invention also provides a non-temporary computer-readable storage medium. When the instructions in the storage medium are executed by a processor of an electronic device, the electronic device can execute and implement a multi-dimensional data-driven automatic strategy mining method as provided above.

[0123] Since the electronic device introduced in this embodiment is an electronic device used to implement the information processing method in the embodiment of the present invention, based on the information processing method introduced in the embodiment of the present invention, a person skilled in the art can understand the specific implementation of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of the present invention is not described in detail here. As long as the electronic device used by a person skilled in the art to implement the information processing method in the embodiment of the present invention, it belongs to the scope of protection of the present invention.

[0124] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0125] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0126] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0127] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0128] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0129] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An automatic strategy mining method based on multidimensional data drive, characterized in that: The method comprises: Obtain feature data sets, label data sets, evaluation index data sets, and actual decision-making result data sets in the financial field within the target time period; Splicing the feature data set, the label data set, the evaluation index data set, and the actual decision result data set to obtain a spliced ​​data set, and dividing the spliced ​​data set into a training set, a test set, and a validation set; Determine an objective function, wherein the objective function is used to determine an optimization direction and an optimization target; Constructing a plurality of decision trees according to the training set, and determining the hit rate and recall rate of each branch in each decision tree, wherein one branch corresponds to one strategy; Based on a preset hit rate threshold, the hit rate and recall rate corresponding to each branch, the corresponding branches are screened to obtain several optimal branches, and the data of the optimal branches in the training set are eliminated to obtain a target strategy set; Evaluating and screening the target strategy set based on the test set and the validation set; Determining weights of a plurality of preset evaluation indicators according to the spliced ​​data set; According to a number of preset evaluation indicators and the weights corresponding to the preset evaluation indicators, the score of each strategy in the target strategy set is determined, and the optimal strategy is obtained.

2. The multidimensional data-driven automatic strategy mining method according to claim 1, characterized in that: Based on a preset hit rate threshold, the hit rate and recall rate corresponding to each branch, the corresponding branches are screened to obtain several optimal branches, and the data of the optimal branches in the training set are removed to obtain a target strategy set, including steps S151-S152: Step S151, based on a preset hit rate threshold, the hit rate and recall rate corresponding to each branch, the corresponding branch is screened to determine whether the branch is the optimal branch; Step S152: if the branch is the optimal branch, remove the data of the branch in the training set and include the branch in the target strategy set; if it is not the optimal branch, re-execute step S151 with other branches.

3. The multi-dimensional data driven automatic strategy mining method according to claim 2, characterized in that: Based on the preset hit rate threshold, the hit rate and recall rate of each branch, the corresponding branches are screened, including: Where SelectBranches(T,θ,N) is the screening function, T is the decision tree, b and b′ are both branches in the decision tree T, HitRate(b) is the hit rate of branch (b), Recall(b) is the recall rate of branch (b), Recall(b′) is the recall rate of branch (b′), θ is the preset hit rate threshold, and N is the number of branches.

4. The automatic strategy mining method based on multidimensional data drive according to claim 1, characterized in that: Objective functions include: Objective Function=maxR(Parameters) HitRate<θ Among them, Objective Function is the objective function, R (Parameters) is the recall rate as a function of the strategy, Parameters is the strategy set, HitRate is the hit rate, and θ is the preset hit rate threshold.

5. The multi-dimensional data driven automatic strategy mining method according to claim 1, characterized in that: The spliced ​​data set is divided into a training set, a test set and a validation set, including: Dividing the target time period into a first time period and a second time period, wherein the time period length of the first time period is greater than the time period length of the second time period and the first time period is before the second time period; Dividing the spliced ​​data in the first time period into a training set and a test set according to a preset ratio; The concatenated data in the second time period is divided into a validation set.

6. The multi-dimensional data driven automatic strategy mining method according to claim 1, characterized in that: Get the optimal strategy, including: The strategy corresponding to the highest score is regarded as the optimal strategy.

7. The multi-dimensional data driven automatic strategy mining method according to claim 1, characterized in that: Determining a score for each strategy in the target strategy set includes: Among them, F(s k ) is strategy s k The score, w i is the weight of the i-th evaluation index, x ki For strategies k The i-th evaluation index of .

8. A strategy automatic mining device driven by multidimensional data, characterized in that: The device comprises: The acquisition module is used to obtain the feature data set, label data set, evaluation index data set and actual decision result data set in the financial field within the target time period; A partitioning module, used to splice the feature data set, the label data set, the evaluation index data set and the actual decision result data set to obtain a spliced ​​data set, and divide the spliced ​​data set into a training set, a test set and a validation set; A function determination module, used to determine an objective function, wherein the objective function is used to determine an optimization direction and an optimization target; A decision module, used to construct a plurality of decision trees according to the training set, and determine the hit rate and recall rate of each branch in each decision tree, wherein one branch corresponds to one strategy; An optimization module, for screening corresponding branches based on a preset hit rate threshold, a hit rate and a recall rate corresponding to each branch, obtaining several optimal branches, and removing data of the optimal branches in the training set to obtain a target strategy set; A screening module, used for evaluating and screening the target strategy set based on the test set and the validation set; A weight module, used to determine the weights of a number of preset evaluation indicators according to the spliced ​​data set; The optimal strategy module is used to determine the score of each strategy in the target strategy set according to a number of preset evaluation indicators and the weights corresponding to each preset evaluation indicator, and obtain the optimal strategy.

9. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute to implement the multi-dimensional data driven automatic strategy mining method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium, characterized in that: When the instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to implement the multidimensional data-driven automatic strategy mining method as described in any one of claims 1 to 7.

Images