Loan credit granting determination method and device, equipment, medium and program product

Abstract

The embodiment of the invention provides a loan credit extension determination method and device, equipment, a medium and a program product, and relates to the field of financial science and technology, the technical field of artificial intelligence or other related fields. The method comprises the following steps: acquiring product planting data; the product planting data is preprocessed and then input into a yield prediction model, the yield corresponding to the product is obtained, and the yield prediction model is obtained by inputting a first training sample set determined based on a target characteristic variable set into a gating circulation unit for training; the target feature variable set is determined based on the influence degree of multiple feature variables in a product planting data sample on the yield; determining a reference score based on the operating parameters and the customer parameters, the operating parameters and the customer parameters being used to indicate the repayment ability of the customer; and determining the credit line and the corresponding interest rate of the customer according to the yield and the reference score. According to the method provided by the invention, the accuracy and objectivity of determining the credit line and the corresponding interest rate are improved.

Description

Technical Field

[0001] The present application relates to the field of financial technology, artificial intelligence technology or other related fields, and in particular to a method, device, equipment, medium and program product for determining loan credit. Background Art

[0002] With the continuous development and popularization of Internet technology, Internet finance has become one of the important focuses of the financial industry. Internet finance has provided new ideas and tools for credit financing, and promoted the development, innovation and efficiency improvement of credit financing products.

[0003] At present, on-site due diligence, credit limit determination and interest rate determination for loan financing are mostly carried out manually. This traditional model occupies an extremely critical position in the credit financing process and fully understands the actual situation of the borrower.

[0004] However, there are a series of problems in conducting on-site due diligence, determining credit limits and interest rates manually, including the influence of subjective factors, high costs, slow information collection, and untimely information updates. Summary of the invention

[0005] The present application provides a method, device, equipment, medium and program product for determining loan credit, which is used to solve the technical problems that manual determination of credit limit and interest rate is affected by subjective factors, has high cost, slow information collection speed and untimely information update.

[0006] In a first aspect, the present application provides a method for determining loan credit, comprising:

[0007] Obtain product planting data;

[0008] The product planting data is pre-processed and input into a yield prediction model to obtain the yield corresponding to the product, wherein the yield prediction model is obtained by inputting a first training sample set determined based on a target feature variable set into a gated recurrent unit for training, wherein the target feature variable set is determined based on the degree of influence of multiple feature variables in the product planting data sample on the yield;

[0009] Determining a reference score based on business parameters and customer parameters, wherein the business parameters and customer parameters are used to indicate the customer's repayment ability;

[0010] The credit limit and corresponding interest rate of the customer are determined based on the output and reference score.

[0011] In a second aspect, the present application provides a device for determining loan credit, comprising:

[0012] Acquisition module, used to obtain product planting data;

[0013] A processing model, used for inputting the product planting data into a yield prediction model after preprocessing to obtain the yield corresponding to the product, wherein the yield prediction model is obtained by inputting a first training sample set determined based on a target feature variable set into a gated recurrent unit for training, wherein the target feature variable set is determined based on the degree of influence of multiple feature variables in the product planting data sample on the yield;

[0014] The processing module is further used to determine a reference score based on business parameters and customer parameters, wherein the business parameters and customer parameters are used to indicate the repayment ability of the customer;

[0015] A determination module is used to determine the credit limit and corresponding interest rate of the customer based on the output and the reference score.

[0016] In a third aspect, the present application provides an electronic device, comprising: a processor, and a memory communicatively connected to the processor;

[0017] The memory stores computer-executable instructions;

[0018] The processor executes the computer-executable instructions stored in the memory to implement the method of the first aspect as described above.

[0019] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the method of the first aspect as described above.

[0020] In a fifth aspect, the present application provides a computer program product, including a computer program, which implements the method of the first aspect when executed by a processor.

[0021] The loan credit determination method, device, equipment, medium and program product provided in this application obtain product planting data through satellite remote sensing technology; then obtain the estimated result of yield through the yield prediction model obtained through gated cyclic unit training, and determine the credit line and corresponding interest rate based on the estimated result of yield, thereby improving the accuracy and objectivity of the determination of the credit line and the corresponding interest rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0023] Figure 1 A schematic diagram of a loan credit determination method provided in an embodiment of the present application Figure 1 ;

[0024] Figure 2 A schematic diagram of a loan credit determination method provided in an embodiment of the present application Figure 2 ;

[0025] Figure 3 A flow chart of a method for determining rock tea loan credit provided in an embodiment of the present application;

[0026] Figure 4 A schematic diagram of a loan credit determination method provided in an embodiment of the present application Figure 3 ;

[0027] Figure 5 A schematic diagram of the structure of a device for determining loan credit provided in an embodiment of the present application;

[0028] Figure 6 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0029] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0030] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0031] 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 application are all information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with the relevant laws, regulations and standards of the relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0032] It should be noted that the loan credit determination method, device, equipment, medium and program product provided in this application can be used in the field of financial technology, artificial intelligence technology or other related fields, and can also be used in any field other than the field of financial technology or other related fields. The application field of the loan credit determination method, device, equipment, medium and program product in this application is not limited.

[0033] As Internet technology continues to expand and become more widespread, Internet finance has become a key area of ​​the financial industry that has attracted much attention. With its unique advantages, Internet finance has opened up new paths and diversified tools for credit financing, and promoted the in-depth development of credit financing products, positive changes in innovation, and significant improvements in efficiency.

[0034] At present, most on-site due diligence of credit financing still relies on manual means. This traditional model occupies a pivotal position in the credit financing process. Its significance lies in being able to gain a more comprehensive insight into the real situation of borrowers and provide first-hand basis for credit decisions, including credit line determination and interest rate determination.

[0035] However, there are also many problems with manual on-site due diligence. From a subjective perspective, subjective factors such as the personal experience and cognitive preferences of investigators affect the objectivity and accuracy of credit limits and interest rate determination. In terms of cost, both labor costs and investigation costs make due diligence costs high.

[0036] The loan credit determination method provided in this application collects product planting data through satellite remote sensing technology and predicts the yield through a yield prediction model, thereby solving the problems of high cost and insufficient objectivity and accuracy in determining the credit limit and interest rate.

[0037] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0038] Figure 1 A schematic diagram of a loan credit determination method provided in an embodiment of the present application Figure 1 .like Figure 1 As shown, the method includes:

[0039] S101. Obtain product planting data;

[0040] Among them, product planting data is used to refer to a series of relevant information collected during the planting process of crops or other products.

[0041] Product planting data is obtained through satellite remote sensing technology. Product planting data may include, for example, planting environment temperature, planting soil elements, planting area, planting variety, geographical location, labor, climate conditions, seedling quantity, pest conditions, growth conditions, etc.

[0042] S102, inputting the product planting data into a yield prediction model after preprocessing to obtain the yield corresponding to the product;

[0043] The yield prediction model is obtained by inputting the first training sample set determined based on the target feature variable set into the gated recurrent unit for training. The target feature variable set is determined based on the influence of multiple feature variables in the product planting data sample on the yield, and is used to indicate features with strong correlation with the yield. The preprocessing includes normalization and standardization processing and feature extraction processing.

[0044] By normalizing and standardizing the product planting data, feature extraction is performed on the normalized and standardized product planting data according to the target feature variable set, the feature vector required by the prediction model is screened out, and the feature vector is input into the yield prediction model to obtain the corresponding yield of the product.

[0045] S103. Determine a reference score based on the business parameters and customer parameters;

[0046] Among them, operating parameters and customer parameters are used to indicate the customer's repayment ability, which are obtained through relevant materials submitted by the customer; operating parameters include product operating scale; method of obtaining operating rights, such as leasing and / or self-ownership; customer's operating experience, such as the number of years of operation of the product; customer parameters include customer age, business location, etc.

[0047] The operating parameters and customer parameters are standardized, and a reference score is determined based on preset weights, where the upper limit of the reference score may be, for example, 100; the preset weights may be adjusted based on the product type; and the standardized processing may include, for example, scoring and normalization.

[0048] The reference score is determined using the following formula:

[0049]

[0050] Among them, V is the reference score; R is the product operation scale; G is the way to obtain the operation right; N is the operation and management; A is the customer age; L is the business location; ω i is the weight.

[0051] Optionally, the reference score can also be adjusted through credit parameters, such as credit honor, etc.

[0052] Exemplary: Table 1 is a reference score table provided in this embodiment. When the product grown by the customer is A tea, the reference score can be determined in the following manner:

[0053] According to Table 1, the management rights of tea mountain A, the method of obtaining the management rights of tea mountain, the number of years engaged in tea business (years), the age of customers and the business location are standardized.

[0054] Table 1

[0055]

[0056]

[0057] Reference score = tea mountain management rights owned (mu) × 40% + tea mountain management rights acquisition method × 30% + years of tea business (years) × 10% + customer age × 10% + business location × 10%

[0058] After the reference score of the customer is determined by the above method, the reference score may be 80, for example, and the reference score of 80 of the customer is adjusted according to the credit parameter.

[0059] S104. Determine the credit limit and corresponding interest rate of the customer based on the output and reference score.

[0060] Optionally, after the loan is issued according to the credit line and the corresponding interest rate, the output of the product can be monitored in real time based on the post-loan product planting data to obtain real-time monitoring results.

[0061] Among them, real-time monitoring results are used for duration management.

[0062] By monitoring product output in real time based on post-loan product planting data, we monitor product output changes in real time, providing a basis for staff to grasp loan quality and risk status in a timely manner.

[0063] The method for determining loan credit provided in this embodiment obtains product planting data through satellite remote sensing technology, thereby improving the objectivity and accuracy of due diligence results; obtains an estimated yield result through a yield prediction model obtained through gated recurrent unit training, better captures long-term dependencies in the data, and improves the accuracy of yield prediction; determines a reference score through operating parameters and customer parameters to provide a reference basis for the credit line and the corresponding interest rate, and determines the credit line and the corresponding interest rate based on the yield and the reference score, thereby improving the accuracy and objectivity of the credit line and the corresponding interest rate.

[0064] Figure 2 A schematic diagram of a loan credit determination method provided in an embodiment of the present application Figure 2 In this embodiment, Figure 1 Based on the embodiment, the method for determining loan credit is described in detail, such as Figure 2 As shown, the method comprises:

[0065] S201, obtaining product planting data;

[0066] S202, inputting the product planting data into a yield prediction model after preprocessing to obtain the yield corresponding to the product;

[0067] S203, determining a reference score based on the business parameters and customer parameters;

[0068] Step S201 , step S202 , and step S203 are similar to step S101 , step S102 , and step S103 , respectively, and are not described again herein.

[0069] S204, estimating the annual sales revenue of the customer based on the output and the unit price corresponding to the product;

[0070] Use the following formula to estimate the customer's annual sales revenue:

[0071]

[0072] Among them, H is the estimated annual sales revenue; W is the output of different varieties of products, P is the market unit price of different varieties of products, and N is the number of varieties of products planted.

[0073] S205, determining an adjustment coefficient according to the reference score and a preset rule;

[0074] Among them, the adjustment coefficient is used to adjust the credit limit.

[0075] It is understandable that the higher the reference score, the stronger the user's repayment ability and the corresponding increase in the credit limit.

[0076] Exemplarily, Table 2 is a table of the relationship between reference scores and adjustment coefficients provided in this embodiment. According to Table 2, the adjustment coefficient is determined:

[0077] Table 2

[0078]

[0079] Exemplary: The adjustment coefficient may also be determined according to the following method:

[0080] M=0.01×V

[0081] S206. Determine a first credit line according to the maximum limit of a single household;

[0082] Among them, the first credit line L 1The maximum limit per household is determined with the credit department. For example, it can be 2 million;

[0083] S207. Calculate the second credit line based on the annual sales income and the adjustment coefficient;

[0084] The second credit limit is determined using the following formula:

[0085] L 2 =H×M×a

[0086] Among them, L 2 is the second credit limit, M is the adjustment coefficient, and a is the credit limit coefficient, which can be 0.5, for example.

[0087] S208. Calculate the third credit line based on the customer's product leased planting area and self-owned planting area;

[0088] The third credit limit is calculated using the following formula:

[0089] L 3 = (C × N 1 +D×N 2 )×ME

[0090] Among them, L 3 is the third credit line, C is the owned planting area, N 1 is the estimated profit coefficient corresponding to the owned planting area; D is the leased planting area, N 2 is the profit coefficient corresponding to the leased planting area; E is the customer's existing credit limit.

[0091] Exemplary: When the product is A tea, N 1 is 50000; N 2 is 20000.

[0092] S209, determining a minimum credit limit from the first credit limit, the second credit limit, and the third credit limit, and using the minimum credit limit as the credit limit of the customer;

[0093] The following formula is used to determine the customer's credit limit:

[0094] L=min[L 1 , L 2 , L 3 ]

[0095] S210: Input the credit line, reference score and annual sales revenue into an interest rate determination model to determine the interest rate corresponding to the credit line.

[0096] Among them, the interest rate determination model is trained by minimizing the difference between the model's interest rate prediction value and the true value of the interest rate.

[0097] Optionally, before determining the interest rate corresponding to the credit limit, it is also necessary to construct an interest rate determination model. The specific process is as follows: obtain a second training sample set; construct an interest rate determination model based on the training feature vector, and calculate the model prediction value based on the interest rate determination model; optimize the interest rate determination model based on the difference between the model prediction value and the true value of the interest rate until the interest rate determination model meets the preset requirements.

[0098] The second training sample set includes a training feature vector, and the training feature vector includes a credit limit, a reference score, an annual sales income, and a true value of an interest rate used for training.

[0099] According to the training feature vector in the training feature vector, an interest rate determination model is constructed based on the support vector machine regression algorithm (SVR), and the interest rate determination model is optimized based on the objective function. The objective function L is expressed as follows:

[0100]

[0101] Where w is the weight vector, b is the bias term, C is the regularization parameter, ||w||^2 is the L2 norm of the weight vector, is the value predicted by the interest rate determination model.

[0102] The first term of the objective function represents minimizing the L2 norm of the weight vector to prevent overfitting. The second term is a square loss function that minimizes the sum of squared errors between the predicted value and the true value. C is a regularization parameter that is used to balance the trade-off between minimizing the error and minimizing the L2 norm of the weight vector.

[0103] There are constraints in the optimization of SVR. The real value of the interest rate of the second training sample set With the model prediction value The difference between needs to be less than or equal to ε, that is , ε is a given tolerance. This constraint can be expressed as:

[0104]

[0105] The interest rate determination model is trained by minimizing the difference between the model's interest rate prediction value and the actual interest rate value, which improves the objectivity and accuracy of interest rate determination.

[0106] For example: Figure 3 The following is a flow chart of a method for determining a rock tea loan credit provided in an embodiment of the present application. Figure 3As shown in the figure, the cloud platform database includes the relevant material information of the customer's application for "Rock Tea Loan" and the rock tea planting data collected frequently and uploaded monthly. According to the rock tea planting data uploaded monthly, the cloud platform database is updated, and the output is obtained based on the pre-processed rock tea planting data and the output prediction model. The annual sales revenue is then estimated based on the output. According to the relevant material information, the reference score is determined, and then the credit line is determined in combination with the estimated annual sales revenue, and then the corresponding interest rate is obtained through the interest rate determination model. The staff can also review the credit line and the corresponding interest rate. In addition, based on the rock tea planting data and the output prediction model uploaded monthly, the rock tea output and estimated annual sales revenue of the customer are monitored in real time.

[0107] The method for determining loan credit provided in this embodiment obtains product planting data through satellite remote sensing technology, thereby improving the objectivity and accuracy of due diligence results; obtains an estimated yield result through a yield prediction model obtained through gated cyclic unit training, better captures long-term dependencies in the data, and improves the accuracy of yield prediction; determines an adjustment coefficient through a reference score, adjusts the credit line, and provides a reference basis for the credit line and the corresponding interest rate; determines the first credit line, the second credit line, and the third credit line respectively, providing a reference basis for determining the credit line, and obtains the interest rate corresponding to the credit line through an interest rate determination model, thereby further improving the accuracy and objectivity of the credit line.

[0108] Figure 4 A schematic diagram of a loan credit determination method provided in an embodiment of the present application Figure 3 In this embodiment, Figure 1 or Figure 2 Based on the embodiment, the process of determining the target feature variable set is described in detail, such as Figure 4 As shown, the method comprises:

[0109] S401, obtaining a product data sample, wherein the product data sample includes output and a plurality of characteristic variables;

[0110] Among them, characteristic variables are used to indicate factors that may affect product yield, including planting environment temperature, planting soil elements, planting area, planting variety, geographical location, labor, climate conditions, seedling quantity, pest conditions, growth conditions, etc.

[0111] The obtained product data sample includes output y and n characteristic variables The product data samples are normalized and standardized to solve the impact of different value ranges and dimensions of each characteristic variable.

[0112] S402, determining a plurality of feature variable sets from the plurality of feature variables;

[0113] The feature variable set includes at least one feature variable.

[0114] The feature variable set includes

[0115] S403, constructing a correlation formula between each characteristic variable set and output;

[0116] The correlation formula is used to indicate the linear regression model constructed according to the correlation between each characteristic variable set and yield.

[0117] The variables in the regression model are the same as the variables in the feature variable set. Based on the a variables in the feature variable set, an a-dimensional regression model is constructed. The formula is as follows:

[0118]

[0119] in, is the output of the a-variable regression model; β is the regression coefficient; x is a variable in the feature variable set.

[0120] Exemplary: Constructing a feature variable set The relationship between the output and the .

[0121] S404: Calculate the significant parameters corresponding to the association equation, and determine the target feature variable from the feature variable set based on the significant parameters and preset parameters, and determine the target feature variable set based on the target feature variable.

[0122] Among them, the significant parameter is used to indicate the degree of influence of the characteristic variable in the correlation equation on the yield; the preset parameter is used to indicate the critical value of the degree of influence of the characteristic variable on the yield.

[0123] The specific process is as follows:

[0124] (1) Introduce the first target feature variable:

[0125] n feature variables , each feature variable is a feature variable set , each characteristic variable set is used to establish a univariate regression model with the dependent variable output y.

[0126]

[0127] The significant parameter corresponding to the correlation equation is calculated using the following formula:

[0128]

[0129] in, When only independent variables are included The regression sum of squares; is the residual sum of squares; n is the sample size.

[0130] According to n characteristic variables, n corresponding correlation equations are constructed, and then the significant parameters corresponding to each correlation equation are calculated, and a maximum value is determined from the n significant parameters. ,Right now:

[0131]

[0132] Set the maximum value The feature variables in the corresponding feature variable set As the target feature variable, the target feature variable is introduced into the target feature variable set.

[0133] (2) Introducing subsequent variables:

[0134] According to the k target feature variables in the target feature variable set, introduce new variables .

[0135] Determine nk feature variable sets:

[0136]

[0137] Understandably, is any feature variable that has not been introduced into the target feature variable set.

[0138] According to the characteristic variable set, a k-dimensional regression model corresponding to the characteristic variable set is constructed, and the corresponding significant parameters are calculated. , the formula is as follows:

[0139]

[0140] in, To include new variables The regression sum of squares of the regression model; To not include new variables The regression sum of squares of the regression model; To not include new variables The residual sum of squares of the regression model.

[0141] After obtaining nk significant parameters, determine a maximum value ,Right now:

[0142]

[0143] judge With the preset threshold The relationship between.

[0144] a) If satisfied , then the new variable in the feature variable set As a target feature variable, and introducing the target feature variable into a target feature variable set;

[0145] b) If not satisfied , then stop introducing new variables , take these k target feature variables as the target feature variable set.

[0146] For example: If the first target feature variable introduced is , it is necessary to introduce the second target feature variable, and then construct the feature variable set as follows:

[0147]

[0148] Based on the above n-1 feature variable sets, a binary regression model is constructed, and the corresponding significant parameters are calculated. Then, the second target feature variable is determined based on the significant parameters, and it is determined whether to introduce the target feature variable into the target feature variable set.

[0149] Optionally, before the product planting data is preprocessed and input into the yield prediction model to obtain the corresponding yield of the product, it is necessary to build a yield prediction model. The specific process is as follows: determine a first training sample set based on the target feature variable set and the training data set; input the first training sample set into a gated recurrent unit, and calculate first information and second information based on the first training sample set and hidden information obtained from the previous training; train the yield prediction model based on the hidden information, the first information and the second information.

[0150] The training data set is obtained in time series; the first information is used to indicate the fusion result of the first training sample set and the hidden information, and the second information is used to indicate the information that needs to be transmitted in the hidden information.

[0151] Gated Recurrent Unit (GRU), the GRU network contains two gates, the update gate and the reset gate. The update gate defines the amount of previous memory saved to the current time step. The reset gate determines how to combine new input information with previous memory. These two gating vectors determine which information can eventually be used as the output of the gated unit. They can save information in long-term sequences and will not be cleared over time or removed because they are not relevant to the prediction.

[0152] According to the target feature variable set and the training data set, the first training sample set is determined, and the features with strong correlation with yield are retained to make the final prediction model more accurate and accelerate the convergence speed of the prediction model.

[0153] The first training sample set is input into the gated recurrent unit.

[0154] The second information is calculated using the following formula:

[0155]

[0156] in, To update the gate, is the activation function; , is the weight matrix of the update gate; To hide information.

[0157] The first information is calculated using the following formula:

[0158]

[0159]

[0160] in, To reset the gate, , is the weight matrix of the reset gate; is the first training sample set; is the first information; W and U are the candidate hidden state weight matrices.

[0161] The new hidden state is obtained using the following formula:

[0162]

[0163] Then, the yield is obtained based on the new hidden state, and the yield prediction model is obtained. For example, the yield can be obtained by adding a fully connected layer after the GRU layer, or by directly outputting the hidden state.

[0164] By capturing long-term dependencies in the data, the accuracy of yield forecasts is improved.

[0165] The method for determining loan credit provided in this embodiment constructs a regression model corresponding to each feature variable set and calculates the corresponding significant parameters, thereby determining the features with strong correlation with yield from the feature variable set, providing a basis for constructing a yield prediction model and improving the accuracy of the yield prediction model.

[0166] Figure 5 A schematic diagram of a loan credit determination device provided in an embodiment of the present application; Figure 5 As shown, the loan credit determination device 50 provided in this embodiment includes:

[0167] An acquisition module 501 is used to acquire product planting data;

[0168] A processing model is used to input the product planting data into the yield prediction model after preprocessing to obtain the yield corresponding to the product, wherein the yield prediction model is obtained by inputting a first training sample set determined based on a target feature variable set into a gated recurrent unit for training, wherein the target feature variable set is determined based on the influence of multiple feature variables in the product planting data sample on the yield;

[0169] The processing module 502 is further used to determine a reference score based on the business parameters and the customer parameters, where the business parameters and the customer parameters are used to indicate the repayment ability of the customer;

[0170] The determination module 503 is used to determine the credit limit and corresponding interest rate of the customer according to the output and the reference score.

[0171] In one possible implementation, the determination module 503 is also used to estimate the customer's annual sales revenue based on the output and the corresponding unit price of the product; determine the adjustment coefficient through the reference score and preset rules, and determine the customer's credit line based on the annual sales revenue and the adjustment coefficient, and the adjustment coefficient is used to adjust the credit line; input the credit line, reference score and annual sales revenue into the interest rate determination model to determine the interest rate corresponding to the credit line, and the interest rate determination model is trained by minimizing the difference between the model's predicted interest rate and the actual interest rate value.

[0172] In one possible implementation, the determination module 503 is also used to determine a first credit line based on the maximum limit for a single household; calculate a second credit line based on annual sales revenue and an adjustment coefficient; calculate a third credit line based on the customer's product leased planting area and owned planting area; determine a minimum credit line from the first credit line, the second credit line, and the third credit line, and use the minimum credit line as the customer's credit line.

[0173] In one possible implementation, the processing module 502 is also used to obtain a second training sample set, the second training sample set includes a training feature vector, the training features include a credit limit, a reference score, an annual sales revenue, and an actual interest rate value used for training; based on the training feature vector, an interest rate determination model is constructed, and a model prediction value is calculated based on the interest rate determination model; based on the difference between the model prediction value and the actual interest rate value, the interest rate determination model is optimized until the interest rate determination model meets preset requirements.

[0174] In one possible implementation, the processing module 502 is also used to obtain a product data sample, the product data sample includes output and multiple characteristic variables; determine multiple characteristic variable sets from the multiple characteristic variables, the characteristic variable sets include at least one characteristic variable; construct a correlation between each characteristic variable set and output; calculate the significant parameters corresponding to the correlation, and based on the significant parameters and preset parameters, determine the target characteristic variables from the characteristic variable set, and determine the target characteristic variable set based on the target characteristic variables, the significant parameters are used to indicate the degree of influence of the characteristic variables in the correlation on the output, and the preset parameters are used to indicate the critical value of the degree of influence of the characteristic variables on the output.

[0175] In a possible implementation, the processing module 502 is also used to determine a first training sample set based on a target feature variable set and a training data set, where the training data set is acquired in a time series; the first training sample set is input into a gated recurrent unit, and first information and second information are calculated based on the first training sample set and hidden information obtained in a previous training, where the first information is used to indicate a fusion result of the first training sample set and the hidden information, and the second information is used to indicate information that needs to be transmitted in the hidden information; and a yield prediction model is trained based on the hidden information, the first information, and the second information.

[0176] In one possible implementation, the product planting data includes post-loan product planting data, and the processing module 502 is also used to monitor the product output in real time based on the post-loan product planting data to obtain real-time monitoring results, which are used for duration management.

[0177] The device for determining loan credit provided in this embodiment can execute the method provided in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be described in detail here.

[0178] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 6 As shown, the electronic device 60 provided in this embodiment includes: at least one processor 601 and a memory 602. Optionally, the device 60 also includes a communication component 603. The processor 601, the memory 602 and the communication component 603 are connected via a bus 604.

[0179] In a specific implementation process, at least one processor 601 executes the computer execution instructions stored in the memory 602, so that at least one processor 601 executes the above method.

[0180] The specific implementation process of the processor 601 can be found in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0181] The present application also provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0182] The present application also provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the above method is implemented.

[0183] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present application.

[0184] It should be further noted that, although the various steps in the flowchart are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowchart may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-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 portion of the sub-steps or stages of other steps.

[0185] It should be understood that the above-mentioned device embodiments are only illustrative, and the device of the present application can also be implemented in other ways. For example, the division of units / modules in the above-mentioned embodiments is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units, modules or components can be combined, or can be integrated into another system, or some features can be ignored or not executed.

[0186] In addition, unless otherwise specified, each functional unit / module in each embodiment of the present application may be integrated into one unit / module, each unit / module may exist physically separately, or two or more units / modules may be integrated together. The above-mentioned integrated unit / module may be implemented in the form of hardware or in the form of a software program module.

[0187] If the integrated unit / module is implemented in the form of hardware, the hardware may be a digital circuit, an analog circuit, etc. The physical implementation of the hardware structure includes but is not limited to transistors, memristors, etc. Unless otherwise specified, the processor may be any appropriate hardware processor, such as a CPU, a GPU, an FPGA, a DSP, an ASIC, etc. Unless otherwise specified, the storage unit may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory (RRAM), a dynamic random access memory (DRAM), a static random access memory (SRAM), an enhanced dynamic random access memory (EDRAM), a high-bandwidth memory (HBM), a hybrid memory cube (HMC), etc.

[0188] If the integrated unit / module is implemented in the form of a software program module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a memory, including a number of instructions to enable a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0189] In the above embodiments, the description of each embodiment has its own emphasis. For the part not described in detail in a certain embodiment, please refer to the relevant description of other embodiments. The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not 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.

[0190] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0191] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A method for determining loan credit, characterized in that: include: Obtain product planting data; The product planting data is pre-processed and input into a yield prediction model to obtain the yield corresponding to the product, wherein the yield prediction model is obtained by inputting a first training sample set determined based on a target feature variable set into a gated recurrent unit for training, wherein the target feature variable set is determined based on the degree of influence of multiple feature variables in the product planting data sample on the yield; Determining a reference score based on business parameters and customer parameters, wherein the business parameters and customer parameters are used to indicate the customer's repayment ability; The credit limit and corresponding interest rate of the customer are determined based on the output and reference score.

2. The method according to claim 1, characterized in that Determining the credit limit and corresponding interest rate of the customer based on the output and the reference score includes: Estimate the annual sales revenue of the customer based on the output and the corresponding unit price of the product; Determine an adjustment coefficient by using the reference score and a preset rule, and determine the credit limit of the customer based on the annual sales revenue and the adjustment coefficient, wherein the adjustment coefficient is used to adjust the credit limit; The credit limit, reference score and annual sales revenue are input into an interest rate determination model to determine the interest rate corresponding to the credit limit. The interest rate determination model is trained by minimizing the difference between the model's interest rate prediction value and the actual interest rate value.

3. The method according to claim 2, characterized in that The determining of the credit limit of the customer based on the annual sales income and the adjustment coefficient includes: Determine the first credit limit based on the maximum limit for a single household; Calculate the second credit line based on the annual sales revenue and adjustment factor; Calculate the third credit line based on the customer's product leased planting area and self-owned planting area; A minimum credit limit is determined from the first credit limit, the second credit limit and the third credit limit, and the minimum credit limit is used as the credit limit of the customer.

4. The method according to claim 2, characterized in that: Before inputting the credit line, reference score and annual sales revenue into an interest rate determination model to determine the interest rate corresponding to the credit line, the method further includes: Acquire a second training sample set, wherein the second training sample set includes a training feature vector, wherein the training feature includes a credit limit, a reference score, an annual sales income, and a true value of an interest rate used for training; constructing an interest rate determination model according to the training feature vector, and calculating a model prediction value according to the interest rate determination model; The interest rate determination model is optimized according to the difference between the model prediction value and the actual value of the interest rate until the interest rate determination model meets the preset requirements.

5. The method according to claim 1, characterized in that Before the product planting data is pre-processed and input into the yield prediction model to obtain the yield corresponding to the product, the method further includes: Acquire a product data sample, wherein the product data sample includes output and a plurality of characteristic variables; Determine a plurality of feature variable sets from the plurality of feature variables, the feature variable sets including at least one feature variable; Construct the correlation between each characteristic variable set and output; Calculate the significant parameters corresponding to the association equation, and determine the target characteristic variables from the characteristic variable set based on the significant parameters and preset parameters, and determine the target characteristic variable set based on the target characteristic variables, the significant parameters are used to indicate the degree of influence of the characteristic variables in the association equation on the yield, and the preset parameters are used to indicate the critical value of the degree of influence of the characteristic variables on the yield.

6. The method according to claim 1, characterized in that Before the product planting data is pre-processed and input into the yield prediction model to obtain the yield corresponding to the product, the method further includes: Determining a first training sample set according to the target feature variable set and the training data set, wherein the training data set is obtained in time series; Inputting the first training sample set into a gated recurrent unit, and calculating first information and second information based on the first training sample set and hidden information obtained in a previous training, wherein the first information is used to indicate a fusion result of the first training sample set and the hidden information, and the second information is used to indicate information to be transmitted in the hidden information; Based on the hidden information, the first information and the second information, a yield prediction model is trained.

7. The method according to claim 1, characterized in that The product planting data includes post-loan product planting data, and the method further includes: based on the post-loan product planting data, real-time monitoring of the output of the product is performed to obtain real-time monitoring results, and the real-time monitoring results are used for duration management.

8. A device for determining loan credit, characterized in that: include: Acquisition module, used to obtain product planting data; A processing model, used for inputting the product planting data into a yield prediction model after preprocessing to obtain the yield corresponding to the product, wherein the yield prediction model is obtained by inputting a first training sample set determined based on a target feature variable set into a gated recurrent unit for training, wherein the target feature variable set is determined based on the degree of influence of multiple feature variables in the product planting data sample on the yield; The processing module is further used to determine a reference score based on business parameters and customer parameters, wherein the business parameters and customer parameters are used to indicate the repayment ability of the customer; A determination module is used to determine the credit limit and corresponding interest rate of the customer based on the output and the reference score.

9. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

11. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 7 when being executed by a processor.

Images