基于遗传算法的模糊C-均值聚类在电信用户行为上的研究

The Research of Fuzzy C-Means Clustering Algorithm Based on Genetic Algorithm in the Telephone User's Behavior

电信用户行为分析,就是指在获得电信用户基本消费数据的情况下,对相关数据进行整理、统计和分析,从中发现用户消费的规律,并将这些规律与电信营销策略等相结合,从而发现目前营销活动中可能存在的问题,并为科学化决策、针对化营销和交叉销售方案的设计提供依据。基于目标函数的模糊C—均值聚类分析作为数据挖掘方法中处理数据的一种,是一种非常优秀的基于划分的模糊集理论分析算法。由于FCM聚类分析在处理数据时容易陷入局部最小值,而遗传算法具有全局寻优的能力。所以将遗传算法与FCM聚类分析相结合,并应用到电信用户行为分析当中,是非常合理的。本文将将遗传算法与FCM聚类分析相结合,并用MATLAB来分析算法的可行性。

Abstract Telecom user behavior analysis, is that in the case of access to telecommunications users basic consumption data, collate relevant data, sum up and analyze the data, which find that the laws of user consumption, and combine these laws and telecommunications marketing strategies, which find that there may be a problem in the current marketing activities, and provide the basis for the design of the scientific decision-maHng, marketing and cross-selling program. Fuzzy C-means clustering analysis is a data mining,is an fuaay analytical and theoretical method based on soft partition. Although the FCM clustering results susceptible initial cluster centers, is easy to fall into local minimum, the genetic algorithm has the capability of global optimization. So the combination of genetic algorithm and FCM apply to the telecommunications user behavior analysis, and use MATLAB to analyze the algorithm.