Power customer comprehensive rating method based data excavation

Abstract

The invention discloses a method for comprehensive credit rating of electric power customers based on data mining, which includes step 1: data collection; step 2: preprocessing the data, extracting relevant data factors and reducing dimensionality; step 3: reducing The time-series statistical factors, kernel distribution factors, and multi-dimensional operation factors in the post-dimensional data factors are automatically extracted, combined and changed, and constructed into new data factors; Step 4: sort out customer-related characteristics, and predict and judge customer behavior in the future; Step 5 : Sort out the three-level indicators of customers, calculate credit scores, and divide credit ratings; solve the lack of marketing professional definitions in the naming and definition of user characteristics in the existing technology credit evaluation system, the calculation method is opaque, and the characteristics are too few to truly reflect the various unique behaviors of users. There is a lack of user characteristics to calculate future trends, and the user boundary is too vague to be integrated into the marketing business, so there is little guidance for the marketing department and other technical problems.

Description

technical field [0001] The invention relates to the field of electric power technology, in particular to a method for comprehensive rating of electric power customers based on data mining. Background technique [0002] With the deepening reform of the electric power system, the groups of electric power customers are more and more complex, and the credit of customers is not the same. The electric power enterprises urgently need an evaluation method to judge the credit of customers. [0003] In the traditional credit evaluation system, the naming definition of user characteristics lacks a marketing professional definition, the calculation method is opaque, and the characteristics are too few to truly reflect the various unique behaviors of users. There is a lack of user characteristics for calculating future trends. When calculating specific user groups The method is relatively simple, and the effect is poor when dealing with highly correlated user characteristics and attribut...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the deficiencies of the above-mentioned prior art, the present invention provides a comprehensive rating method for electric power customers based on data mining, to solve the lack of marketing professional definition in the naming definition of user features in the credit evaluation system of the prior art, the opaque calculation method, and the feature Too few can not truly reflect the various unique behaviors of users, and there is a lack of user characteristics for calculating future trends. The method is relatively simple when calculating specific user groups, and the effect is poor when processing highly correlated user characteristic attributes. The user boundary is too vague and difficult to integrate into the marketing business, so it is of little help to the marketing department's guidance and other technical issues

Images