Special installment credit line adjustment method and device, equipment and storage medium

Abstract

The invention relates to the technical field of financial science and technology, in particular to a special installment credit line adjustment method, device and equipment and a storage medium, and the method comprises the steps: processing customer credit data based on a target random forest model to obtain an initial credit line; obtaining additional credit data in response to the condition that the initial credit limit is not less than a preset limit threshold value; processing the initial credit line and the additional credit data based on a preset amount adjustment proportion calculation model to obtain a target amount adjustment proportion; and adjusting the initial credit line based on the target amount adjustment proportion to obtain a target credit line. The method is convenient for improving the credit line adjustment effect when a bank examines and approves the special installment application of the customer.

Description

Technical Field

[0001] The present application relates to the field of financial technology, and in particular to a method, device, equipment and storage medium for adjusting a special installment credit limit. Background Art

[0002] Bank special installment is an installment loan service provided by the bank based on the customer's credit status to meet the customer's specific consumption purposes (such as home improvement, car purchase, education, travel, etc.). It is based on the customer's credit limit, and the consumption amount is divided into several installments for repayment within a certain period, and each special project usually has its own specific limit range, interest rate, term and other regulations.

[0003] Before obtaining the qualification for special installment, customers need to apply to the bank for a special installment credit line. Currently, when banks approve customers' special installment applications based on their credit status, they generally adjust the credit line applied for by customers. However, since the bank's process for approving special installment applications based on customers' credit status is redundant, and the special installment account adjustment process is not independent and the rules are not granular enough, the accuracy of the credit line adjustment is poor. It can be seen that the bank's adjustment effect of the credit line when approving customers' special installment applications is poor. Summary of the invention

[0004] In order to facilitate improving the credit limit adjustment effect of the bank when approving a customer's special installment application, the present application provides a special installment credit limit adjustment method, device, equipment and storage medium.

[0005] In a first aspect, the present application provides a special installment credit limit adjustment method, comprising:

[0006] Process customer credit data based on the target random forest model to obtain the initial credit limit;

[0007] In response to the initial credit limit being not less than a preset credit limit threshold, acquiring additional credit data;

[0008] Processing the initial credit limit and the additional credit data based on a preset credit limit adjustment ratio calculation model to obtain a target credit limit adjustment ratio;

[0009] The initial credit limit is adjusted based on the target credit limit adjustment ratio to obtain a target credit limit.

[0010] Through the above implementation, when the customer's credit data and additional credit data are obtained, the adjusted target credit limit can be calculated according to the target random forest model and the adjustment ratio calculation model, avoiding the problems of process redundancy and non-independence of the special installment account adjustment process and insufficient rule granularity. This makes it easier to improve the credit limit adjustment effect of the bank when approving the customer's special installment application.

[0011] Preferably, the step of constructing the target random forest model includes:

[0012] Dividing the training sample data into a plurality of decision sample data based on the decision rule;

[0013] Constructing a one-to-one corresponding target decision tree based on the decision sample data;

[0014] Determine the initial weight of each target decision tree respectively, and iteratively adjust the initial weight based on the decision sample data corresponding to the target decision tree to obtain the target weight;

[0015] A target random forest model is obtained based on each of the target decision trees and the target weights corresponding to each of the target decision trees.

[0016] Through the above implementation, the constructed target random forest model facilitates the subsequent calculation of the initial credit limit.

[0017] Preferably, the constructing a one-to-one corresponding target decision tree based on the decision sample data includes:

[0018] Processing the decision sample data based on a preset extraction ratio to obtain a decision training set and a decision test set;

[0019] Constructing a corresponding initial decision tree based on the decision training set;

[0020] The target decision tree is obtained by pruning the initial decision tree based on the decision test set.

[0021] Through the above implementation, it is easy to prevent the target random forest model from overfitting.

[0022] Preferably, the iterative adjustment of the initial weights based on the decision sample data corresponding to the target decision tree to obtain the target weights includes:

[0023] Processing the decision sample data based on a preset extraction ratio to obtain a decision verification set;

[0024] The initial weight is iteratively optimized based on the decision verification set until a preset iteration stop condition is met to obtain a target weight.

[0025] Through the above implementation, it is easy to improve the accuracy of the initial credit limit calculated subsequently.

[0026] Preferably, the step of processing customer credit data based on the target random forest model to obtain an initial credit limit includes:

[0027] Based on each target decision tree in the target random forest model, customer credit data is processed respectively to obtain a one-to-one corresponding decision amount;

[0028] The initial credit limit is obtained by weighting each decision amount based on the target weight corresponding to each target decision tree.

[0029] Through the above implementation, it is easy to calculate an accurate initial credit limit.

[0030] Preferably, the above-mentioned special installment credit limit adjustment method further includes:

[0031] In response to the initial credit limit being less than a preset limit threshold, the business of applying for the special installment payment is stopped.

[0032] Through the above implementation, it is easy to determine whether it is necessary to continue to apply for special installment services for the customer based on the initial credit limit and the preset credit limit threshold.

[0033] In a second aspect, the present application provides a special installment credit limit adjustment device, comprising:

[0034] A model processing module is used to process customer credit data based on a target random forest model to obtain an initial credit limit;

[0035] A data acquisition module, configured to acquire additional credit data in response to the initial credit limit being not less than a preset credit limit threshold;

[0036] A ratio calculation module, configured to process the initial credit limit and the additional credit data based on a preset credit limit ratio calculation model to obtain a target credit limit ratio;

[0037] The credit limit adjustment module is used to adjust the initial credit limit based on the target credit limit adjustment ratio to obtain a target credit limit.

[0038] Through the above implementation, when the customer's credit data and additional credit data are obtained, the adjusted target credit limit can be calculated according to the target random forest model and the adjustment ratio calculation model, avoiding the problems of process redundancy and non-independence of the special installment account adjustment process and insufficient rule granularity. This makes it easier to improve the credit limit adjustment effect of the bank when approving the customer's special installment application.

[0039] In a third aspect, the present application provides a computer device, the computer device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the steps in the above method when executing the computer program.

[0040] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which implements the steps in the above method when executed by a processor.

[0041] In a fifth aspect, the present application further provides a computer program product, wherein the computer program product comprises a computer program, and when the computer program is executed by a processor, the steps in any of the above method embodiments are implemented.

[0042] The above-mentioned special installment credit limit adjustment method, device, equipment and storage medium obtain an initial credit limit by processing customer credit data based on a target random forest model; in response to the initial credit limit being not less than a preset limit threshold, obtain additional credit data; process the initial credit limit and the additional credit data based on a preset adjustment ratio calculation model to obtain a target adjustment ratio; adjust the initial credit limit based on the target adjustment ratio to obtain a target credit limit. Through the above implementation, when the customer credit data and additional credit data are obtained, the adjusted target credit limit can be calculated according to the target random forest model and the adjustment ratio calculation model, avoiding the problems of process redundancy, non-independence of the special installment account adjustment process, and insufficient rule granularity, so as to facilitate improving the credit limit adjustment effect of the bank when approving the customer's special installment application.

[0043] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 A flow chart of a special installment credit limit adjustment method provided in an embodiment of the present application;

[0046] Figure 2 This is a structural schematic diagram of a special installment credit limit adjustment device provided in an embodiment of the present application;

[0047] Figure 3 A schematic diagram of the structure of a computer device provided in an embodiment of the present application;

[0048] Figure 4 This is a diagram of the internal structure of a computer-readable storage medium provided in an embodiment of the present application. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present disclosure more clear, the present disclosure is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present disclosure and are not used to limit the present disclosure.

[0050] It should be noted that the terms "first", "second", etc. in the specification and claims of this article and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of this article described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or equipment that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0051] In this article, the term "and / or" is only a description of the association relationship between related 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 related objects before and after are in an "or" relationship.

[0052] Embodiment 1

[0053] Figure 1 A flow chart of a special installment credit limit adjustment method provided in Example 1 of this application, refer to Figure 1 The method may be performed by a device for performing the method, and the device may be implemented by software and / or hardware. The method includes:

[0054] S110. Process customer credit data based on the target random forest model to obtain an initial credit limit.

[0055] Among them, the target random forest model is a random forest model that has been trained. The target random forest model is used to process the customer credit data input therein and output the corresponding initial credit limit. The customer credit data is the financial data required to evaluate the customer's credit level when the customer applies for a switch to installment payment. Exemplarily, the customer credit data includes: the total amount of previous special installment payment amounts, the arrears score, the average annual income in the past five years, the number of one-time default entries, the utilization rate of the credit limit at the end of the bill, the amount of installment blacklist hits in consumption, and the amount of outstanding installments, etc. The initial credit limit is the credit limit preliminarily calculated by the target random forest model based on the customer credit data. However, since the actual and accurate credit limit that meets the customer's credit level is also affected by other factors, the initial credit limit will need to be adjusted later.

[0056] Specifically, the customer's credit data is obtained, and then the customer credit data is input into a trained target random forest model for processing, and the target random forest model outputs an initial credit limit.

[0057] S120: In response to the initial credit limit being not less than a preset credit limit threshold, obtaining additional credit data.

[0058] Among them, the credit limit threshold is a preset value set according to historical experience data. The credit limit threshold is used to compare with the initial credit limit, so as to determine whether to continue to apply for special installments for the customer in the future; if the initial credit limit is less than the credit limit threshold, it means that the customer's credit level is not good, and there is no need to continue to apply for special installments for the customer in the future; on the contrary, if the initial credit limit is not less than the credit limit threshold, it means that the customer's credit level is good; special installments can be continued to be applied for the customer in the future; the above-mentioned actual and accurate credit limit that meets the customer's credit level is also affected by other factors, and the other factors are also the additional credit data mentioned in this step; illustratively, the additional credit data in this embodiment includes: the proportion of early installment repayments, the proportion of large installment amounts, and the proportion of credit card overdue times, etc. In other embodiments, there is no specific limitation.

[0059] It should be noted that when the initial credit limit is not less than the preset limit threshold, obtaining additional credit data facilitates further adjustment of the initial credit limit based on the additional credit data.

[0060] Specifically, when the initial credit limit is not less than a preset limit threshold, additional credit data is obtained.

[0061] S130. Processing the initial credit limit and the additional credit data based on a preset credit limit adjustment ratio calculation model to obtain a target credit limit adjustment ratio.

[0062] Among them, the credit adjustment ratio calculation model is a calculation model that has been trained. The credit adjustment ratio calculation model is used to process the initial credit limit and additional credit data input therein, and then output the target credit adjustment ratio. In this embodiment, the credit adjustment ratio calculation model is a calculation model obtained by training a deep convolutional neural network. The credit adjustment ratio calculation model is provided with: a convolution layer, a downsampling layer, a fully connected layer and an output layer, and a Softmax function is provided in the output layer; when calculating the initial credit limit and additional credit data, the credit adjustment ratio calculation model performs at least two rounds of convolution calculation and downsampling calculation on the initial credit limit and additional credit data in sequence based on the convolution layer and the downsampling layer therein, thereby obtaining the corresponding downsampling result, and then processes the downsampling result through the fully connected layer to obtain a single value; further, the single value is processed by the Softmax function in the output layer to obtain the target credit adjustment ratio. The target credit adjustment ratio represents the adjustment ratio when the initial credit limit is adjusted later. For example, the target credit adjustment ratio is +0.3, -0.3, etc.

[0063] It should be noted that, in the present embodiment, the adjustment ratio calculation model is a model that has been trained. During the training process of the adjustment ratio calculation model: the corresponding training data will be divided into training set, test set and validation set according to the ratio of 6:2:2. Since the input is manually selected, the selected indicators may affect the final result. Therefore, a simple cross-validation method is adopted to discard part of the input and select the model with the best performance on the validation set as the adjustment ratio calculation model. In addition, during the training process, the random gradient descent optimization method is used to adjust the feedforward parameters of the model according to the performance results. This process is repeated until the training ends after the loss converges. Since the adjustment ratio calculation model uses a convolutional neural network, data regression prediction may produce oversampling, which may lead to overfitting of the model. Therefore, the model will be pruned and neurons at different levels will be discarded (dropout).

[0064] Specifically, the initial credit limit and the additional credit data are input into a preset credit limit adjustment ratio calculation model for processing, and the credit limit adjustment ratio calculation model outputs a target credit limit adjustment ratio.

[0065] S140: Adjust the initial credit limit based on the target credit limit adjustment ratio to obtain a target credit limit.

[0066] The target credit limit is the credit limit obtained by adjusting the initial credit limit by the target adjustment ratio; for example, assuming that the initial credit limit is M and the target adjustment ratio is 0.3, the target credit limit is M+0.3M=1.3M.

[0067] Specifically, the initial credit limit is adjusted by the target adjustment ratio to obtain the target credit limit.

[0068] It should be noted that this embodiment processes customer credit data based on a target random forest model to obtain an initial credit limit; in response to the initial credit limit being not less than a preset credit limit threshold, obtains additional credit data; processes the initial credit limit and the additional credit data based on a preset credit limit adjustment ratio calculation model to obtain a target credit limit adjustment ratio; and adjusts the initial credit limit based on the target credit limit adjustment ratio to obtain a target credit limit. Through the above implementation, when customer credit data and additional credit data are obtained, the adjusted target credit limit can be calculated based on the target random forest model and the credit limit adjustment ratio calculation model, avoiding the problems of process redundancy, non-independence of the special installment account credit limit adjustment process, and insufficient rule granularity, and improving the accuracy of the target credit limit finally calculated, which facilitates improving the credit limit adjustment effect of the bank when approving a customer's special installment application.

[0069] It should also be noted that the method provided in this embodiment speeds up the approval process, improves the accuracy of the results and reduces labor costs; the innovative training method and decision tree construction method make the model more accurate with limited data, saving the bank's labor costs and enhancing customer experience.

[0070] Embodiment 2

[0071] A special installment credit limit adjustment method is provided in Example 2 of the present application. The method refines the "target random forest model construction step" in Example 1. It should be noted that for the parts not described in detail in this embodiment, reference can be made to the descriptions of other embodiments. The method includes:

[0072] S210. Process customer credit data based on the target random forest model to obtain an initial credit limit.

[0073] Among them, the steps of constructing the target random forest model include:

[0074] A210. Divide the training sample data into a plurality of decision sample data based on the decision rule.

[0075] Among them, in this embodiment, the training sample data includes historical data on the total amount of special installment applications in the past, historical data on the arrears score, historical data on the average annual income in the past five years, historical data on the number of one-time default deposits, historical data on the usage rate of the amount at the end of the bill, historical data on the amount of installment blacklist hits in consumption, and historical data on the amount of outstanding installments; the decision rules are used to divide the above-mentioned training sample data to obtain multiple decision sample data, and each decision sample data is used to subsequently construct a corresponding decision tree; the decision rules can combine different types of historical data in the training sample data to achieve the division of the training sample data.

[0076] Specifically, the training sample data is divided into a plurality of decision sample data according to the decision rule.

[0077] A220. Construct a one-to-one target decision tree based on the decision sample data.

[0078] Among them, each decision sample data is used to construct a one-to-one corresponding decision tree; the target decision tree is the corresponding decision tree constructed by the decision sample data.

[0079] Specifically, according to each decision sample data, a target decision tree corresponding to each decision sample data is constructed.

[0080] A230. Determine the initial weights of each target decision tree respectively, and iteratively adjust the initial weights based on the decision sample data corresponding to the target decision tree to obtain the target weights.

[0081] Among them, the decision rules can divide different historical data combinations from the training sample data; each historical data combination is also a decision sample data, and the decision sample data is subsequently used to calculate the corresponding target decision tree; the target random forest model contains multiple target decision trees, and the multiple target decision trees subsequently need to perform regression operations on customer credit data, so it is necessary to determine the weight of each target decision tree; the weight of the target decision tree is initially determined by the historical data combination corresponding to the target decision tree, and the weight of the corresponding target decision tree initially determined by the historical data combination is recorded as the initial weight.

[0082] It should be noted that the initial weights cannot accurately represent the true weights of the corresponding target decision trees. Therefore, the weights of each target decision tree need to be optimized later. In this embodiment, the decision sample data corresponding to the target decision tree can be used to iteratively optimize the initial weights. The result obtained after iterative optimization of the initial weights is the target weight N, which can reflect the true weight of the corresponding target decision tree.

[0083] Specifically, the initial weight corresponding to the target decision tree is preliminarily determined based on the historical data combination corresponding to the target decision tree; further, the initial weight is iteratively optimized through the decision sample data corresponding to the target decision tree, so as to obtain the target weight N corresponding to the initial weight.

[0084] It should be noted that the decision sample data used to construct each target decision tree are different, which helps to reduce the overfitting problem of the random forest model to a certain extent.

[0085] A240. Obtain a target random forest model based on each of the target decision trees and the target weights corresponding to each of the target decision trees.

[0086] Among them, multiple target decision trees with target weights corresponding to each other can together constitute a target random forest model.

[0087] Specifically, multiple target decision trees with target weights corresponding to each other are combined into a target random forest model.

[0088] S220: In response to the initial credit limit being not less than a preset credit limit threshold, obtaining additional credit data.

[0089] S230: Process the initial credit limit and the additional credit data based on a preset credit limit adjustment ratio calculation model to obtain a target credit limit adjustment ratio.

[0090] S240: Adjust the initial credit limit based on the target credit limit adjustment ratio to obtain a target credit limit.

[0091] Embodiment 3

[0092] A special installment credit limit adjustment method is provided in Example 3 of the present application. This method refines the "constructing a one-to-one corresponding target decision tree based on the decision sample data" in Example 2. It should be noted that for the parts not described in detail in this embodiment, reference can be made to the descriptions of other embodiments. This method includes:

[0093] S310. Process customer credit data based on the target random forest model to obtain an initial credit limit.

[0094] Among them, the steps of constructing the target random forest model include:

[0095] A310. Divide the training sample data into a plurality of decision sample data based on the decision rule.

[0096] A321. Process the decision sample data based on a preset extraction ratio to obtain a decision training set and a decision test set.

[0097] Among them, the extraction ratio is a preset ratio set according to historical experience data; exemplarily, the extraction ratio is A:B:C, which characterizes the proportional relationship between the training set, the test set and the validation set; in this embodiment, the extraction ratio between the training set and the test set is A:B=1:2, and the extraction ratio is used to divide each decision sample data separately, so as to divide a single decision sample data into a decision training set, a decision test set and a decision validation set.

[0098] Specifically, the decision sample data is divided by a preset extraction ratio to obtain a decision training set and a decision test set.

[0099] A322. Construct a corresponding initial decision tree based on the decision training set.

[0100] The decision training set is used to construct an initial decision tree corresponding to the decision sample data that divides the decision training set.

[0101] Specifically, taking one of the decision sample data as an example, a decision training set divided by the decision sample data is obtained; and then an initial decision tree corresponding to the decision sample data is constructed through the decision training set.

[0102] A323. Prune the initial decision tree based on the decision test set to obtain a target decision tree.

[0103] It should be noted that the initial decision tree initially established based on the decision test set may have too high a fit to the decision training set, and almost remembers every detail in the decision training set, resulting in poor performance when facing new test data; in order to solve the overfitting problem, in this embodiment, the decision test set can be divided into decision sample data to perform post-pruning on the initial decision tree, and the decision tree obtained after post-pruning the initial decision tree is also the target decision tree; illustratively, each node is evaluated step by step from the leaf node of the initial decision tree upward, for each node, assuming that the node and the branches below it are removed, the error rate of the decision tree is re-evaluated using the decision test set, if the error rate does not increase or increases very little (within an acceptable range) after removing the node, then the node and the branches below it are removed, thereby achieving post-pruning of the initial decision tree; in other embodiments, there is no specific limitation.

[0104] Specifically, the decision test set is divided into decision sample data to perform post-pruning on the initial decision tree, thereby obtaining the target decision tree.

[0105] A330. Determine the initial weights of each target decision tree respectively, and iteratively adjust the initial weights based on the decision sample data corresponding to the target decision tree to obtain the target weights.

[0106] A340. Obtain a target random forest model based on each of the target decision trees and the target weights corresponding to each of the target decision trees.

[0107] S320: In response to the initial credit limit being not less than a preset credit limit threshold, obtaining additional credit data.

[0108] S330: Process the initial credit limit and the additional credit data based on a preset credit limit adjustment ratio calculation model to obtain a target credit limit adjustment ratio.

[0109] S340: Adjust the initial credit limit based on the target credit limit adjustment ratio to obtain a target credit limit.

[0110] Embodiment 4

[0111] A special installment credit limit adjustment method is provided in Example 4 of the present application. This method refines the "iterative adjustment of the initial weight to obtain the target weight based on the decision sample data corresponding to the target decision tree" in Example 2; it should be noted that for the parts not described in detail in this embodiment, reference can be made to the descriptions of other embodiments. This method includes:

[0112] S410. Process customer credit data based on the target random forest model to obtain an initial credit limit.

[0113] Among them, the steps of constructing the target random forest model include:

[0114] A410. Divide the training sample data into a plurality of decision sample data based on the decision rule.

[0115] A420. Construct a one-to-one target decision tree based on the decision sample data.

[0116] A431. Determine the initial weights of each of the target decision trees.

[0117] A432. Process the decision sample data based on a preset extraction ratio to obtain a decision verification set.

[0118] Among them, the decision sample data is divided according to a preset extraction ratio, and a decision verification set corresponding to the decision sample data can also be generated.

[0119] Specifically, the decision sample data is divided according to a preset extraction ratio to obtain a decision verification set.

[0120] A433. Iteratively optimize the initial weight based on the decision verification set until a preset iteration stop condition is met to obtain a target weight.

[0121] Among them, the decision verification set is used to iteratively optimize the initial weights of the target decision tree corresponding to the decision sample data; taking the first optimization as an example, the decision verification set adjusts the initial weights to the weights after the first adjustment, and then normalizes the weights after the first adjustment to obtain the weights after the first optimization; then the weights after the first optimization are optimized for the second time according to the above steps; in this embodiment, the iterative stopping condition is that the number of iterative optimizations reaches a preset optimization number threshold, and in other embodiments, there is no specific limitation; the target weight is the iterative optimization weight corresponding to the preset iterative stopping condition.

[0122] Specifically, the initial weights are iteratively optimized through the decision verification set until the preset iteration stop condition is met, thereby obtaining the target weight.

[0123] A440. Obtain a target random forest model based on each of the target decision trees and the target weights corresponding to each of the target decision trees.

[0124] S420: In response to the initial credit limit being not less than a preset credit limit threshold, obtaining additional credit data.

[0125] S430: Process the initial credit limit and the additional credit data based on a preset credit limit adjustment ratio calculation model to obtain a target credit limit adjustment ratio.

[0126] S440: Adjust the initial credit limit based on the target credit limit adjustment ratio to obtain a target credit limit.

[0127] Embodiment 5

[0128] A special installment credit limit adjustment method is provided in Example 5 of the present application. This method refines the "processing customer credit data based on the target random forest model to obtain an initial credit limit" in Example 1. It should be noted that for the parts not described in detail in this embodiment, please refer to the description of other embodiments. The method includes:

[0129] S511. Based on each target decision tree in the target random forest model, the customer credit data is processed respectively to obtain a one-to-one corresponding decision amount.

[0130] Among them, the target random forest model contains multiple target decision trees, which are used to process customer credit data and output the decision amount corresponding to the target decision tree.

[0131] Specifically, the acquired customer credit data is input into multiple target decision trees in the target random forest model for processing, thereby obtaining the decision amounts output by each target decision tree.

[0132] S512: Perform weighted calculation on each decision amount based on the target weight corresponding to each target decision tree to obtain an initial credit limit.

[0133] Among them, each target decision tree has its corresponding target weight; each target weight can be used to perform weighted calculation on the corresponding decision amounts to obtain an initial credit limit.

[0134] Specifically, the target weight corresponding to each target decision tree and the decision amount output by each target decision tree are obtained; then, each decision amount is weightedly calculated by the obtained target weights to obtain the initial credit limit.

[0135] S520: In response to the initial credit limit being not less than a preset credit limit threshold, obtaining additional credit data.

[0136] S530: Process the initial credit limit and the additional credit data based on a preset credit limit adjustment ratio calculation model to obtain a target credit limit adjustment ratio.

[0137] S540: Adjust the initial credit limit based on the target credit limit adjustment ratio to obtain a target credit limit.

[0138] Embodiment 6

[0139] A special installment credit limit adjustment method provided in Example 6 of the present application can be executed by a device for executing the method, and the device can be implemented by software and / or hardware. The method includes:

[0140] S610: Process customer credit data based on the target random forest model to obtain an initial credit limit.

[0141] S620: In response to the initial credit limit being not less than a preset credit limit threshold, obtaining additional credit data.

[0142] S630: Process the initial credit limit and the additional credit data based on a preset credit limit adjustment ratio calculation model to obtain a target credit limit adjustment ratio.

[0143] S640: Adjust the initial credit limit based on the target credit limit adjustment ratio to obtain a target credit limit.

[0144] S650: In response to the initial credit limit being less than a preset credit limit threshold, stop applying for the special installment service.

[0145] Among them, if the initial credit limit is less than the preset limit threshold, it means that the credit level of the customer corresponding to the initial credit limit is not good. At this time, it is generally believed that the customer is not allowed to apply for special installments, so the business of applying for special installments for the customer needs to be stopped at this time.

[0146] Specifically, when the initial credit limit is less than the preset limit threshold, the special installment business for the corresponding customer will be stopped.

[0147] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0148] Embodiment 7

[0149] Based on the same inventive concept, this embodiment also provides a special installment credit limit adjustment device for implementing the special installment credit limit adjustment method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more special installment credit limit adjustment device embodiments provided below can refer to the limitations of the special installment credit limit adjustment method above, and will not be repeated here.

[0150] In this embodiment, Figure 2 As shown, a special installment credit limit adjustment device is provided, comprising:

[0151] A model processing module is used to process customer credit data based on a target random forest model to obtain an initial credit limit;

[0152] A data acquisition module, configured to acquire additional credit data in response to the initial credit limit being not less than a preset credit limit threshold;

[0153] A ratio calculation module, configured to process the initial credit limit and the additional credit data based on a preset credit limit ratio calculation model to obtain a target credit limit ratio;

[0154] The credit limit adjustment module is used to adjust the initial credit limit based on the target credit limit adjustment ratio to obtain a target credit limit.

[0155] Each module in the above-mentioned special installment credit limit adjustment device can be implemented in whole or in part by software, hardware and their combination. Each module can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0156] It should be noted that this embodiment processes customer credit data based on a target random forest model to obtain an initial credit limit; in response to the initial credit limit being not less than a preset credit limit threshold, obtains additional credit data; processes the initial credit limit and the additional credit data based on a preset credit limit adjustment ratio calculation model to obtain a target credit limit adjustment ratio; and adjusts the initial credit limit based on the target credit limit adjustment ratio to obtain a target credit limit. Through the above implementation, when customer credit data and additional credit data are obtained, the adjusted target credit limit can be calculated based on the target random forest model and the credit limit adjustment ratio calculation model, avoiding the problems of process redundancy, non-independence of the special installment account credit limit adjustment process, and insufficient rule granularity, and improving the accuracy of the target credit limit finally calculated, which facilitates improving the credit limit adjustment effect of the bank when approving a customer's special installment application.

[0157] Embodiment 8

[0158] In this embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a special installment credit limit adjustment method is implemented.

[0159] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the scheme of the present disclosure, and does not constitute a limitation on the computer device to which the scheme of the present disclosure is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0160] Embodiment 9

[0161] In this embodiment, a computer readable storage medium is provided. Figure 4 As shown, a computer program is stored thereon, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0162] Embodiment 10

[0163] In this embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0164] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties.

[0165] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided by the present disclosure can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided by the present disclosure may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited thereto. The processor involved in each embodiment provided by the present disclosure may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited thereto.

[0166] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0167] The above-described embodiments only express several implementation methods of the present disclosure, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present disclosure, and these all belong to the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be subject to the attached claims.

Claims

1. A special installment credit limit adjustment method, characterized in that: include: Process customer credit data based on the target random forest model to obtain the initial credit limit; In response to the initial credit limit being not less than a preset credit limit threshold, acquiring additional credit data; Processing the initial credit limit and the additional credit data based on a preset credit limit adjustment ratio calculation model to obtain a target credit limit adjustment ratio; The initial credit limit is adjusted based on the target credit limit adjustment ratio to obtain a target credit limit.

2. The method according to claim 1, characterized in that The steps of constructing the target random forest model include: Dividing the training sample data into a plurality of decision sample data based on the decision rule; Constructing a one-to-one corresponding target decision tree based on the decision sample data; Determine the initial weight of each target decision tree respectively, and iteratively adjust the initial weight based on the decision sample data corresponding to the target decision tree to obtain the target weight; A target random forest model is obtained based on each of the target decision trees and the target weights corresponding to each of the target decision trees.

3. The method according to claim 2, characterized in that The constructing a one-to-one corresponding target decision tree based on the decision sample data includes: Processing the decision sample data based on a preset extraction ratio to obtain a decision training set and a decision test set; Constructing a corresponding initial decision tree based on the decision training set; The target decision tree is obtained by pruning the initial decision tree based on the decision test set.

4. The method according to claim 2, characterized in that: The iteratively adjusting the initial weight based on the decision sample data corresponding to the target decision tree to obtain the target weight includes: Processing the decision sample data based on a preset extraction ratio to obtain a decision verification set; The initial weight is iteratively optimized based on the decision verification set until a preset iteration stop condition is met to obtain a target weight.

5. The method according to claim 1, characterized in that The step of processing customer credit data based on the target random forest model to obtain an initial credit limit includes: Based on each target decision tree in the target random forest model, customer credit data is processed respectively to obtain a one-to-one corresponding decision amount; The initial credit limit is obtained by weighting each decision amount based on the target weight corresponding to each target decision tree.

6. The method according to claim 1, characterized in that Also includes: In response to the initial credit limit being less than a preset limit threshold, the business of applying for the special installment payment is stopped.

7. A special installment credit limit adjustment device, characterized in that: The device comprises: A model processing module is used to process customer credit data based on a target random forest model to obtain an initial credit limit; A data acquisition module, configured to acquire additional credit data in response to the initial credit limit being not less than a preset credit limit threshold; A ratio calculation module, configured to process the initial credit limit and the additional credit data based on a preset credit limit ratio calculation model to obtain a target credit limit ratio; The credit limit adjustment module is used to adjust the initial credit limit based on the target credit limit adjustment ratio to obtain a target credit limit.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images