Credit pre-granting and measuring system based on multi-layer perceptron algorithm
The pre-credit credit measurement system based on the multi-layer perceptron algorithm solves the problems of low efficiency and difficulty in controlling the accuracy of traditional credit assessment methods, achieves efficient and accurate pre-credit credit measurement calculations, and reduces the probability of calculation errors.
Patent Information
- Application Number
- CN202411361601.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-27
- Publication Date
- 2025-01-21
AI Technical Summary
Traditional credit assessment methods are inefficient and error-prone, and cannot effectively and accurately evaluate customer credit risks, making it difficult to guarantee the efficiency and accuracy of pre-credit credit measurement.
A pre-credit credit measurement system based on the multi-layer perceptron algorithm is adopted to realize automated pre-credit credit measurement calculation through data collection, pre-processing, multi-layer perceptron algorithm calculation and result output modules. The system uses the backpropagation algorithm to train a multi-layer perceptron model, optimizes model performance, and obtains multi-data source information through distributed data acquisition technology.
It has significantly improved the efficiency and accuracy of pre-credit amount calculation, and can accurately and efficiently calculate the pre-credit amount, and through continuous training and adjustment, the probability of calculation errors has been reduced.
Smart Images
Figure CN119338574A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of financial technology, and specifically is a pre-credit limit measurement system based on a multi-layer perceptron algorithm. Background Art
[0002] In the financial field, it is crucial for banks and financial institutions to accurately assess customer credit risk and reasonably set pre-credit limits. Traditional credit assessment methods usually rely on manual review and limited data analysis, which is not only inefficient but also difficult to control in terms of accuracy. With the rise of big data and machine learning technology, a series of financial models have become feasible to achieve automated calculations with the help of algorithms, thus opening up a new solution for efficient pre-credit measurement.
[0003] Existing solution business personnel must use data obtained from multiple channels to manually calculate and analyze credit, risk, region and other data to arrive at the final pre-credit limit, which is inefficient and prone to errors. Summary of the invention
[0004] In view of the above problems, the present invention is proposed to provide a pre-credit limit measurement system based on a multi-layer perceptron algorithm that overcomes the above problems or at least partially solves the above problems.
[0005] To achieve the above object, the present invention adopts the following technical solutions:
[0006] A pre-credit limit measurement system based on a multi-layer perceptron algorithm, the system comprising:
[0007] Data collection module, used to collect data related to pre-credit limit measurement;
[0008] A data preprocessing module is used to preprocess the collected data;
[0009] Multilayer perceptron algorithm module, used to obtain the final pre-credit limit measurement result;
[0010] The credit limit measurement result output module is used to output the pre-credit limit measurement results.
[0011] Optionally, the system further includes a model evaluation and optimization module, which is used to evaluate the performance of the multilayer perceptron algorithm model and to adjust and improve the multilayer perceptron algorithm model using an optimization algorithm.
[0012] Optionally, the data collection module uses distributed data collection technology to establish connections with multiple data sources to obtain data related to pre-credit limit measurement in real time.
[0013] Optionally, the data related to the pre-credit limit measurement includes the user's basic information, credit record, financial status and tax information.
[0014] Optionally, the data preprocessing module uses a data cleaning algorithm to remove invalid and abnormal data, and uses a data normalization algorithm to map the data.
[0015] Optional, multilayer perceptron algorithm module includes:
[0016] Input layer unit, receiving preprocessed data;
[0017] Hidden layer units perform nonlinear transformation and feature extraction on data through neurons;
[0018] The output layer unit outputs the final pre-credit limit measurement result.
[0019] Optionally, the multilayer perceptron algorithm module constructs a multilayer perceptron network structure including multiple hidden layer units, and trains the multilayer perceptron algorithm model using a back propagation algorithm.
[0020] Optionally, the credit limit measurement result output module organizes and formats the pre-credit limit measurement results, and transmits the pre-credit limit measurement values and key indicators in the calculation process to relevant personnel.
[0021] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:
[0022] The present invention is based on a multi-layer perceptron algorithm and has powerful prediction and regression analysis capabilities. It can significantly improve the efficiency of pre-credit limit calculation and accurately and efficiently calculate the pre-credit limit. Moreover, through continuous training and adjustment, the probability of calculation errors can be greatly reduced. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 A schematic diagram of a pre-credit limit measurement system based on a multi-layer perceptron algorithm provided in an embodiment of the present application;
[0024] Figure 2 A schematic diagram of the architecture of a pre-credit limit measurement system based on a multi-layer perceptron algorithm provided in an embodiment of the present application. DETAILED DESCRIPTION
[0025] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0026] See also Figure 1 and Figure 2This embodiment provides a pre-credit limit measurement system based on a multi-layer perceptron algorithm, the system comprising:
[0027] The data collection module is used to collect data related to pre-credit limit measurement.
[0028] The data related to the pre-credit limit measurement includes the user's basic information, credit record, financial status and tax information, etc.
[0029] The data collection module uses distributed data collection technology to establish connections with multiple data sources to obtain data related to pre-credit measurement in real time. Tax data and other data need to be obtained from third-party APIs and stored in the database for backup.
[0030] The data preprocessing module is used to preprocess the collected data.
[0031] The collected data is preprocessed by cleaning, conversion, and normalization. After processing, the suitable data is input into the multi-layer perceptron algorithm for training and prediction.
[0032] The data preprocessing module uses data cleaning algorithms to remove invalid and abnormal data, and uses data normalization algorithms to map the data. These preprocessed data vectors are passed to the neurons in the input layer, and each neuron uses the received data as an input signal.
[0033] The multi-layer perceptron algorithm module is used to obtain the final pre-credit limit measurement results.
[0034] The multilayer perceptron algorithm module includes:
[0035] Input layer unit receives preprocessed data.
[0036] Hidden layer units perform nonlinear transformation and feature extraction on data through neurons.
[0037] The output layer unit outputs the final pre-credit limit measurement result.
[0038] During the training phase, the multilayer perceptron is trained using historical data and the connection weights between neurons are adjusted to improve prediction accuracy.
[0039] The multi-layer perceptron algorithm module constructs a multi-layer perceptron network structure containing multiple hidden layer units, and uses the back propagation algorithm to train the multi-layer perceptron algorithm model. Among them, the nonlinear mapping ability of the multi-layer perceptron algorithm is used to learn the complex relationship in the data, thereby achieving accurate prediction of the measurement amount; its adaptive learning ability continuously adjusts the network parameters to optimize the model performance.
[0040] The model evaluation and optimization module is used to evaluate the performance of the multilayer perceptron algorithm model, and to adjust and improve the multilayer perceptron algorithm model using an optimization algorithm based on the evaluation results to improve its performance.
[0041] Evaluation methods such as cross-validation are used to measure the accuracy and generalization ability of the model. The feedback adjustment ability of the multi-layer perceptron algorithm is used to adjust the algorithm's hyperparameters, such as learning rate and number of hidden layer nodes, according to the evaluation results to optimize the model.
[0042] The credit limit measurement result output module is used to output the pre-credit limit measurement results.
[0043] The pre-credit limit estimation results calculated by the multi-layer perceptron algorithm are output and presented to relevant business personnel in a clear and intuitive manner.
[0044] The credit limit measurement result output module organizes and formats the pre-credit limit measurement results and sends them to the display system via RPC. The display system displays them in an intuitive and easy-to-understand page and transmits the pre-credit limit measurement values and key indicators in the calculation process to relevant personnel.
[0045] This embodiment is based on a multi-layer perceptron algorithm, and has powerful prediction and regression analysis capabilities. It can significantly improve the efficiency of pre-credit limit calculations and accurately and efficiently calculate pre-credit limits. And through continuous training and adjustment, it can greatly reduce the probability of calculation errors.
[0046] Model Introduction:
[0047] Multi-Layer Perceptron (MLP) is a feedforward artificial neural network model. It usually consists of an input layer, multiple hidden layers, and an output layer. Neurons are connected by weights, and learning and pattern recognition of data are achieved by adjusting these weights.
[0048] Development History:
[0049] The development of multilayer perceptrons can be traced back to the early research on neural networks. With the improvement of computing power and the continuous improvement of algorithms, it has been widely used in pattern recognition, classification, prediction and other fields. In recent years, with the rise of deep learning technology, multilayer perceptrons have also continued to evolve and expand, and combined with other technologies, have achieved better performance.
[0050] advantage:
[0051] Powerful nonlinear mapping capabilities can handle complex nonlinear problems.
[0052] It has good generalization ability and can make relatively accurate predictions on new data after proper training.
[0053] It is highly scalable and can adapt to tasks of different scales and complexities by increasing the number of hidden layers and neurons.
[0054] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.
Claims
1. A pre-credit limit measurement system based on a multi-layer perceptron algorithm, characterized in that: The system comprises: Data collection module, used to collect data related to pre-credit limit measurement; A data preprocessing module is used to preprocess the collected data; Multilayer perceptron algorithm module, used to obtain the final pre-credit limit measurement result; The credit limit measurement result output module is used to output the pre-credit limit measurement results.
2. A pre-credit limit measurement system based on a multi-layer perceptron algorithm as claimed in claim 1, characterized in that: The system also includes a model evaluation and optimization module, which is used to evaluate the performance of the multi-layer perceptron algorithm model and use an optimization algorithm to adjust and improve the multi-layer perceptron algorithm model.
3. A pre-credit limit measurement system based on a multi-layer perceptron algorithm as claimed in claim 1, characterized in that: The data collection module uses distributed data collection technology to establish connections with multiple data sources and obtain data related to pre-credit limit measurement in real time.
4. A pre-credit limit measurement system based on a multi-layer perceptron algorithm as claimed in claim 3, characterized in that: The data related to the pre-credit limit measurement includes the user's basic information, credit record, financial status and tax information.
5. A pre-credit limit measurement system based on a multi-layer perceptron algorithm as claimed in claim 1, characterized in that: The data preprocessing module uses data cleaning algorithms to remove invalid and abnormal data, and uses data normalization algorithms to map the data.
6. A pre-credit limit measurement system based on a multi-layer perceptron algorithm as claimed in claim 1, characterized in that: The multilayer perceptron algorithm module includes: Input layer unit, receiving preprocessed data; Hidden layer units perform nonlinear transformation and feature extraction on data through neurons; The output layer unit outputs the final pre-credit limit measurement result.
7. A pre-credit limit measurement system based on a multi-layer perceptron algorithm as claimed in claim 6, characterized in that: The multilayer perceptron algorithm module constructs a multilayer perceptron network structure containing multiple hidden layer units, and uses the back propagation algorithm to train the multilayer perceptron algorithm model.
8. A pre-credit limit measurement system based on a multi-layer perceptron algorithm as claimed in claim 1, characterized in that: The credit limit measurement result output module organizes and formats the pre-credit limit measurement results, and transmits the pre-credit limit measurement values and key indicators in the calculation process to relevant personnel.