Customer classification method and system and electronic equipment

Abstract

The embodiment of the invention provides a customer classification method and system and electronic equipment. The customer classification method comprises the steps: obtaining a mixed data sample setcomposed of a plurality of mixed data samples, wherein each mixed data sample is composed of a numerical variable sample and a classification variable sample; performing feature extraction on the numerical variable sample to obtain a comprehensive evaluation factor; obtaining a dissimilarity measurement value of the comprehensive evaluation factor and the classification variable sample; and optimizing the K-Means clustering algorithm of the dissimilarity measurement value based on a particle swarm optimization algorithm, constructing a customer classification model, and further obtaining a customer classification result. By combining the K-Means clustering algorithm and the improved particle swarm optimization algorithm, and considering multiple indexes influencing customer consumption, the customer classification method carries out clustering analysis on customer consumption data by adopting an optimized clustering algorithm to obtain customer groups with different characteristics, so that the analysis result is more reasonable and accurate, and different operation and customer service strategies can be conveniently adopted for different groups.

Description

technical field [0001] The embodiment of the present invention relates to the technical field of computer information processing, in particular to a customer classification method, system and electronic equipment. Background technique [0002] Retail customer classification refers to the behavior of dividing customers into different customer groups by using relevant technical analysis methods based on factors such as customer attributes, purchase needs, and values, to evaluate the consumption behavior of different customers, so as to find high-value customers, or according to classification As a result, tailor-made corresponding products and services for customers. The traditional methods of classifying retail customers mainly include experience segmentation method and mathematical statistics method. The methods are all one-sided. [0003] In data mining technology, cluster analysis, as an unsupervised learning method, can classify data sets and discover valuable informati...

AI Technical Summary

Problems solved by technology

Among them, the K-Means clustering algorithm proposed by MacQueen in 1967 is a partition-based clustering method, but the K-Means clustering algorithm also has the following deficiencies: (1) In calculating the three-dimensional relationship between data objects When the degree of heterogeneity is high, the clustering algorithm can only deal with numerical data, not suitable for processing data with types; and it does not balance the importance of different features; (2) The selection of random initial values ​​may lead to large clustering results. Uncertainty, there are even situations where there is no solution and trapped in a local minimum

[0005] To sum up, the methods for classifying retail customers in the prior art have many deficiencies, resulting in inaccurate analysis results and inconspicuous classification effects.

Images