Customer payment habit analysis and prediction method and system

The SMOTE algorithm enhances customer electricity consumption data, combines NILM and Apriori algorithm to analyze customer payment behavior, and combines the habit analysis model to perform oversampling processing, which solves the problem of SOM neural network being sensitive to outliers and improves the stability and accuracy of customer payment habit prediction.

CN119692629AActive Publication Date: 2025-03-25国网安徽省电力有限公司营销服务中心
5 Cites 0 Cited by

Patent Information

Application Number
CN202510199851.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-02-24
Publication Date
2025-03-25
Estimated Expiration
2045-02-24

AI Technical Summary

Technical Problem

In the prior art, SOM neural networks are sensitive to outliers and noise, resulting in large deviations and unstable results of unbalanced customer electricity data clustering.

Method used

The SMOTE algorithm is used to enhance the processing of customer electricity consumption data, and the customer payment behavior portrait is obtained through NILM and Apriori algorithm analysis, and the evaluation index vector is oversampled with the habit analysis model, and the change feature quantity is extracted to predict customer payment habits.

Benefits of technology

It effectively alleviates the problem of habit analysis model being sensitive to outliers and noise, improves the stability and accuracy of clustering results for unbalanced customer electricity consumption data, and avoids overfitting.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a customer payment habit analysis and prediction method and a customer payment habit analysis and prediction system, belongs to the technical field of power operation, and solves the problems of large deviation and instability of an unbalanced customer power consumption data clustering result caused by the fact that an SOM neural network is sensitive to an abnormal value and noise in an existing analysis method. The method comprises the following steps: performing customer electricity consumption data enhancement processing based on an SMOTE algorithm, performing overall oversampling processing on an evaluation index vector by adopting a habit analysis model, calculating an output matrix of an evaluation index, analyzing and processing the output matrix of the evaluation index based on a difference feature extraction method, and outputting a payment habit prediction result corresponding to a customer; according to the method, the habit analysis model is high in customer payment behavior portrait analysis precision and good in robustness, and the problem that the habit analysis model is sensitive to abnormal values and noise can be effectively relieved by enhancing the customer power consumption data based on the SMOTE algorithm.
Need to check novelty before this filing date? Find Prior Art