CN114021612A - Novel personal credit assessment method and system - Google Patents
Novel personal credit assessment method and system Download PDFInfo
- Publication number
- CN114021612A CN114021612A CN202111080243.7A CN202111080243A CN114021612A CN 114021612 A CN114021612 A CN 114021612A CN 202111080243 A CN202111080243 A CN 202111080243A CN 114021612 A CN114021612 A CN 114021612A
- Authority
- CN
- China
- Prior art keywords
- credit
- data
- classification
- types
- credit data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24137—Distances to cluster centroïds
- G06F18/2414—Smoothing the distance, e.g. radial basis function networks [RBFN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Business, Economics & Management (AREA)
- Software Systems (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Marketing (AREA)
- Probability & Statistics with Applications (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Medical Informatics (AREA)
- Development Economics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention relates to a novel personal credit evaluation method and a novel personal credit evaluation system, which mainly comprise the main links of data acquisition, data cleaning, model training and the like, wherein the acquired data are respectively processed from two aspects of data quantity and characteristic value, and three classification models of SVM, GWOO-Sigmoid and GA-BP are integrated. The invention ensures the balance of the sample data of each credit type by data augmentation in the data acquisition stage, eliminates the colinearity among the credit data by means of the correlation coefficient to determine the credit data type participating in credit evaluation, and provides a good data basis for the training of a classification model. In addition, the invention integrates the classification results of the three classification models by utilizing the idea of integrated learning so as to obtain the final credit evaluation result, thereby avoiding the generalization problem of a single model and further improving the accuracy of credit evaluation.
Description
Technical Field
The invention belongs to the technical field of credit assessment, and particularly relates to a novel personal credit assessment method and system.
Background
In the current big data age, internet technology and artificial intelligence technology develop rapidly, and people's living consumption is constantly changing, and changes from the past consumption of saving money to the present consumption of credit. Since the network personal loan service is increasing, personal credit becomes an indispensable profit point for financial service institutions such as commercial banks, and thus, when a financial service institution such as each commercial bank issues credit to an individual, it is necessary to evaluate the credit status of the individual in advance, so as to reduce the threat to the financial service institutions such as commercial banks due to personal credit risk and ensure the income of the financial service institutions in personal credit business.
The personal credit system is a system for recording the credit activities of consumers in detail, and is the basis for constructing a developed credit consumption economy in a social range. The role of the personal credit system as the basis of the social credit system is becoming more and more important.
Credit evaluation is essentially a classification problem in pattern recognition, i.e., a business or individual consumer is classified by credit. The method specifically comprises the steps of finding out the characteristics of each credit category according to a plurality of samples of each category in history, summarizing classification rules, establishing a mathematical model, predicting the credit category of a borrower and providing credit evaluation basis for banks and other financial institutions.
With the increasing market competition and the rapid development of computer technology, more and more methods (such as quantitative analysis tools of statistics and operational research) are applied to the field of credit assessment. The statistical method mainly comprises linear regression, discriminant analysis, logistic regression and the like, and the operation research method mainly refers to some linear programming methods. Most credit scoring models use one or a combination of methods. In recent years, some nonparametric statistical methods and artificial intelligence models have been introduced into scoring models, such as neural networks, expert systems, genetic algorithms, and nearest neighbor methods in nonparametric statistics.
In recent years, deep learning becomes a research topic of fire and heat, and has the obvious advantages that data features can be automatically extracted, deep learning can be understood as a deep neural network with a plurality of nonlinear layers, and a computer can automatically learn mode features and blend the feature learning into the process of establishing a model, so that the incompleteness of the model caused by artificial design features is reduced. Some machine learning applications taking deep learning as a core reach recognition or classification performance exceeding that of the existing algorithm under the application scene meeting specific conditions.
Disclosure of Invention
In order to further improve the accuracy of credit assessment, the invention provides a novel personal credit assessment method and a novel personal credit assessment system, which adopt the following technical scheme:
a novel personal credit assessment method comprises the following steps:
step 1: collecting various types of credit data of the loan clients with the determined credit types;
step 2: preprocessing the collected credit data to remove abnormal data, delete repeated data, eliminate co-linearity among the credit data and determine the credit data types participating in credit evaluation;
and step 3: extracting characteristic variables of credit data participating in credit evaluation through a tree-based machine learning algorithm and a principal component analysis method;
and 4, step 4: constructing a training sample set by using the characteristic variables and credit types of loan clients corresponding to the characteristic variables, and respectively training SVM, GWOO-Sigmoid and GA-BP classification models through the training sample set;
and 5: collecting credit data of a client to be assessed, acquiring the credit data of the client participating in credit assessment, and extracting characteristic variables by the tree-based machine learning algorithm and the principal component analysis method in the step 3;
step 6: inputting the characteristic variables extracted in the step 5 into trained SVM, GWOO-Sigmoid and GA-BP classification models respectively to obtain corresponding classification results, wherein the classification result of each model is the probability that the client to be evaluated belongs to each credit type;
and 7: and weighting and summing the probabilities of the corresponding credit types in the three classification model classification results, wherein the weight is determined by the Kappa coefficient of each classification model, and the credit type with the maximum probability after summation is used as the final credit type of the client to be evaluated.
Further, the credit types include extremely high credit customers, good credit customers, general credit customers, poor credit customers, and extremely low credit customers.
Further, in step 1, in order to ensure the data balance, a value range of the same credit data of the same credit type loan clients is obtained, and a value is randomly selected in the value range as the credit data to increase the data amount of the credit type.
Further, in step 2, if some credit data of some credit type loan clients are missing, filling the credit data with the mean value of the same credit data of the same credit type loan clients; and when the collinearity between the credit data is eliminated, calculating the correlation coefficient between the credit data respectively, if the correlation coefficient of a certain two types of credit data is larger than a certain threshold value, taking the mean value of the two types of credit data as the credit data participating in credit evaluation, and otherwise, reserving the two types of credit data and participating in credit evaluation.
Further, the correlation coefficient is calculated by the formula
Where r represents the correlation coefficient of two types of credit data,
and
means, x, representing two kinds of credit data, respectivelyiAnd yiRespectively representing two kinds of credit data of the ith loan client, wherein n is the total number of the loan clients obtaining the credit data; the correlation threshold valueIs 0.85.
The credit evaluation system based on the credit evaluation method comprises a data acquisition module, a data preprocessing module and a classification module; the data acquisition module is used for acquiring various credit data of a client to be evaluated; the data preprocessing module acquires credit data participating in credit evaluation according to various collected credit data; the classification module comprises an algorithm execution unit and trained SVM, GWO-Sigmoid and GA-BP classification models, wherein the algorithm execution unit is used for executing a tree-based machine learning algorithm and a principal component analysis method in the step 3 to extract characteristic variables of credit data participating in credit evaluation, the characteristic variables are respectively input into the trained SVM, GWO-Sigmoid and GA-BP classification models to obtain corresponding classification results, and finally the final credit types of the clients to be evaluated are obtained and output in a weighted summation mode.
Compared with the prior art, the method ensures the balance of the sample data of each credit type through data augmentation in the data acquisition stage, eliminates the collinearity among the credit data by means of the correlation coefficient to determine the credit data type participating in credit evaluation, and provides a good data basis for training of a classification model. In addition, the invention integrates the classification results of the three classification models by utilizing the idea of integrated learning so as to obtain the final credit evaluation result, thereby avoiding the generalization problem of a single model and further improving the accuracy of credit evaluation.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a schematic diagram of an SVM classification model;
FIG. 3 is a schematic diagram of the GWO-Sigmoid classification model;
FIG. 4 is a schematic diagram of a GA-BP classification model.
Detailed Description
The present invention will now be described in further detail with reference to the accompanying drawings.
As shown in fig. 1, the credit evaluation method of the present invention mainly comprises the following steps:
(1) the method comprises the steps of collecting various types of credit data of loan clients with determined credit types, dividing the loan clients into five types of clients with extremely high credit degree, good credit degree, ordinary credit degree, poor credit degree and extremely poor credit degree, wherein the credit data of the loan clients comprise various types of data such as age, historical loan amount, historical credit level and the like.
(2) Data augmentation and data cleansing of collected data to determine the types of credit data that are involved in credit evaluation.
In fact, the loan clients with low credit degree are far less than the loan clients with high credit degree, the clients with lower credit degree have fewer samples when collecting data, and the training of the classification model has certain requirements on the number of various types of samples, namely, the number of various types of data samples is not much different. Aiming at the problem, the minimum value and the maximum value of each type of credit data value are found out from all samples of a certain type of client, and then a value is randomly selected in a value range to serve as the type of credit data, so that the number of the samples of the client is increased.
For the collected credit data of various clients, whether missing values exist or not is checked firstly, if the missing values exist, the missing values are found out and filled with the mean value of the credit data of the same type of clients, and therefore the influence of the existence of abnormal values on subsequent model training is avoided. And finally, performing collinearity analysis on various credit data, determining whether multiple collinearity exists among the credit data according to the magnitude of a correlation coefficient among the credit data, generally, if the correlation coefficient is large, indicating that multiple collinearity exists among variables, and eliminating the influence of the multiple collinearity on the result by removing one credit data or fusing several credit data with correlation into a new variable (for example, calculating the mean value of several correlation credit data and taking the result as a new credit data type). The correlation coefficient is calculated by
Where r represents the correlation coefficient of two types of credit data,
and
means, x, representing two kinds of credit data, respectivelyiAnd yiRespectively representing two kinds of credit data of the ith loan client, wherein n is the total number of the loan clients acquiring the credit data, and the multiple collinearity is removed when the correlation coefficient is greater than 0.85. .
(3) Feature variables of credit data participating in credit evaluation are extracted. When the characteristic variables are selected, the main characteristic variables are extracted by adopting the idea of combining a machine learning algorithm based on a tree and a principal component analysis method. Variable screening using a tree-based machine learning algorithm is a relatively inexpensive approach. The model itself is used for prediction, but in the prediction process, the importance of the variables can be ranked and then the screening of the variables can be performed by this ranking. The method is suitable for a series of machine learning algorithms based on trees, such as random forests, xgboost, lightgbm and the like, and has a prominent effect in practical application. After some more important variables are screened out, because there may be correlation between the variables, when there is a certain correlation between two variables, it can be understood that there is a certain overlap between the information of the two variables. The principal component analysis is to eliminate redundant repeated variables (closely related variables, which are related to each other after being screened by a tree-based machine learning algorithm) for all the variables, and to establish as few new variables as possible, so that the new variables are unrelated to each other, and the new variables keep original information as much as possible.
(4) The extracted feature variables and credit types of loan clients corresponding to the feature variables are used for constructing a training sample set, SVM, GWO-Sigmoid (Sigmoid neural network improved based on the Grey wolf algorithm) and GA-BP (BP neural network improved based on the genetic algorithm) classification models are respectively trained through the training sample set, and schematic diagrams of the SVM, GWO-Sigmoid and GA-BP classification models are respectively shown in fig. 2, fig. 3 and fig. 4. In the training process, the classification performance of the model can be judged by observing the accuracy, the recall rate and the AUC value of the classification model in real time so as to determine whether to stop training.
(5) And collecting credit data of the client to be assessed, acquiring the credit data participating in credit assessment and extracting the characteristic variable by the same method.
(6) And respectively inputting the extracted characteristic variables into trained SVM, GWOO-Sigmoid and GA-BP classification models to obtain corresponding classification results, wherein the classification result of each model is the probability that the customer to be evaluated belongs to each credit type.
(7) And weighting and summing the probabilities of the corresponding credit types in the three model classification results based on the idea of ensemble learning, wherein the weight is determined by the Kappa coefficient of each classification model, and the credit type with the maximum probability after summation is used as the final credit type of the client to be evaluated.
Based on the credit assessment method, the invention also provides a credit assessment system which comprises a data acquisition module, a data preprocessing module and a classification module. The data pre-processing module acquires credit data participating in credit evaluation according to the acquired various credit data. The classification module comprises an algorithm execution unit and trained SVM, GWOO-Sigmoid and GA-BP classification models, wherein the algorithm execution unit is used for executing a tree-based machine learning algorithm and a principal component analysis method in the step (3) to extract the characteristic variables of credit data participating in credit evaluation, inputting the characteristic variables into the trained SVM, GWOO-Sigmoid and GA-BP classification models respectively to obtain corresponding classification results, and then obtaining and outputting the final credit types of the clients to be evaluated in a weighted summation mode.
The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above-mentioned embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that modifications and embellishments within the scope of the invention may be made by those skilled in the art without departing from the principle of the invention.
Claims (6)
1. A novel personal credit assessment method, comprising the steps of:
step 1: collecting various types of credit data of the loan clients with the determined credit types;
step 2: preprocessing the collected credit data to remove abnormal data, delete repeated data, eliminate co-linearity among the credit data and determine the credit data types participating in credit evaluation;
and step 3: extracting characteristic variables of credit data participating in credit evaluation through a tree-based machine learning algorithm and a principal component analysis method;
and 4, step 4: constructing a training sample set by using the characteristic variables and credit types of loan clients corresponding to the characteristic variables, and respectively training SVM, GWOO-Sigmoid and GA-BP classification models through the training sample set;
and 5: collecting credit data of a client to be assessed, acquiring the credit data of the client participating in credit assessment, and extracting characteristic variables by the tree-based machine learning algorithm and the principal component analysis method in the step 3;
step 6: inputting the characteristic variables extracted in the step 5 into trained SVM, GWOO-Sigmoid and GA-BP classification models respectively to obtain corresponding classification results, wherein the classification result of each model is the probability that the client to be evaluated belongs to each credit type;
and 7: and weighting and summing the probabilities of the corresponding credit types in the three classification model classification results, wherein the weight is determined by the Kappa coefficient of each classification model, and the credit type with the maximum probability after summation is used as the final credit type of the client to be evaluated.
2. The novel personal credit evaluation method of claim 1, wherein said credit types include extremely high credit customers, good credit customers, general credit customers, poor credit customers, and extremely low credit customers.
3. The method as claimed in claim 1, wherein in step 1, in order to ensure the data balance, a value range of the same credit data of the same credit type loan clients is obtained, and a value is randomly selected in the value range as the credit data to increase the data amount of the credit type.
4. The novel personal credit assessment method according to claim 1, wherein in step 2, if some credit data of some credit type loan client is missing, the method is filled with the mean value of the same credit data of the same credit type loan client; and when the collinearity between the credit data is eliminated, calculating the correlation coefficient between the credit data respectively, if the correlation coefficient of certain two types of credit data is larger than a correlation threshold value, taking the mean value of the two types of credit data as the credit data participating in credit evaluation, and otherwise, reserving and participating in credit evaluation.
5. The method of claim 4, wherein the correlation coefficient is calculated by the formula
Where r represents the correlation coefficient of two types of credit data,
and
means, x, representing two kinds of credit data, respectivelyiAnd yiRespectively representing two kinds of credit data of the ith loan client, wherein n is the total number of the loan clients obtaining the credit data; the correlation threshold is 0.85.
6. The credit evaluation system based on the credit evaluation method of any one of claims 1 to 5, comprising a data acquisition module, a data preprocessing module and a classification module;
the data acquisition module is used for acquiring various credit data of a client to be evaluated;
the data preprocessing module acquires credit data participating in credit evaluation according to various collected credit data;
the classification module comprises an algorithm execution unit and trained SVM, GWO-Sigmoid and GA-BP classification models, wherein the algorithm execution unit is used for executing a tree-based machine learning algorithm and a principal component analysis method in the step 3 to extract characteristic variables of credit data participating in credit evaluation, the characteristic variables are respectively input into the trained SVM, GWO-Sigmoid and GA-BP classification models to obtain corresponding classification results, and finally the final credit types of the clients to be evaluated are obtained and output in a weighted summation mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111080243.7A CN114021612A (en) | 2021-09-15 | 2021-09-15 | Novel personal credit assessment method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111080243.7A CN114021612A (en) | 2021-09-15 | 2021-09-15 | Novel personal credit assessment method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114021612A true CN114021612A (en) | 2022-02-08 |
Family
ID=80054139
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111080243.7A Pending CN114021612A (en) | 2021-09-15 | 2021-09-15 | Novel personal credit assessment method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114021612A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116993490A (en) * | 2023-08-15 | 2023-11-03 | 广州佳新智能科技有限公司 | Automatic bank scene processing method and system based on artificial intelligence |
| CN117934159A (en) * | 2024-03-21 | 2024-04-26 | 北京信立合创信息技术有限公司 | Personal credit report query monitoring and early warning method based on artificial intelligence |
-
2021
- 2021-09-15 CN CN202111080243.7A patent/CN114021612A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116993490A (en) * | 2023-08-15 | 2023-11-03 | 广州佳新智能科技有限公司 | Automatic bank scene processing method and system based on artificial intelligence |
| CN116993490B (en) * | 2023-08-15 | 2024-03-01 | 广州佳新智能科技有限公司 | Automatic bank scene processing method and system based on artificial intelligence |
| CN117934159A (en) * | 2024-03-21 | 2024-04-26 | 北京信立合创信息技术有限公司 | Personal credit report query monitoring and early warning method based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bahnsen et al. | Example-dependent cost-sensitive decision trees | |
| Abdou et al. | Neural nets versus conventional techniques in credit scoring in Egyptian banking | |
| CN111311402A (en) | XGboost-based internet financial wind control model | |
| US20160055427A1 (en) | Method for providing data science, artificial intelligence and machine learning as-a-service | |
| US20160078367A1 (en) | Data clean-up method for improving predictive model training | |
| CN110956273A (en) | Credit scoring method and system integrating multiple machine learning models | |
| Sayjadah et al. | Credit card default prediction using machine learning techniques | |
| Tiffin | Machine learning and causality: The impact of financial crises on growth | |
| CN114021612A (en) | Novel personal credit assessment method and system | |
| Chen et al. | Loan default prediction using diversified sensitivity undersampling | |
| CN112529683A (en) | Method and system for evaluating credit risk of customer based on CS-PNN | |
| Qiu et al. | Credit risk scoring analysis based on machine learning models | |
| Wang et al. | A multi-factor two-stage deep integration model for stock price prediction based on intelligent optimization and feature clustering | |
| Chen et al. | A deep learning approach using DeepGBM for credit assessment | |
| Negi et al. | Cryptocurrency Price Analysis using Deep Learning | |
| CN116993490B (en) | Automatic bank scene processing method and system based on artificial intelligence | |
| Tsai et al. | Data pre-processing by genetic algorithms for bankruptcy prediction | |
| He et al. | Self-Adaptive bagging approach to credit rating | |
| Zhou | Data mining for individual consumer credit default prediction under e-commence context: a comparative study | |
| Osowski et al. | Prediction of customer status in corporate banking using neural networks | |
| Jakka et al. | Original Research Article A novel credit scoring system in financial institutions using artificial intelligence technology | |
| Datkhile et al. | Statistical modelling on loan default prediction using different models | |
| CN115496364A (en) | Method and device for identifying heterogeneous enterprises, storage medium and electronic equipment | |
| Wu et al. | Customer churn prediction for commercial banks using customer-value-weighted machine learning models | |
| CN114418604A (en) | Method and device for predicting survival promotion success probability of palm silver easy-to-lose customers and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |