摘要
针对客户服务项目的不确定性,基于不可分辨关系的粗糙集理论和BP神经网络算法优良的分类映射能力,提出了面向细分客户群的基于粗糙BP神经网络客户群特征与服务项目映射模型.本文将分析客户特征,运用粗糙集理论进行客户特征约简、划分等价关系、建立BP神经网络的初始拓扑结构,运用K-means算法划分客户群.通过引入粗糙集理论,改进BP神经网络算法,加快BP网络收敛的速度和逃离局部极小值点,并利用rosetta软件和Matlab编程实现面向细分客户群的客户特征与服务项目映射模型.
Aimed at the uncertainty of customer service project,this paper,which based on rough sets and BP neural network and facing on segmented customer groups,present a mapping model of the customer base characteristics and service project,using the undiscerning relation in rough set theory and the excellent classification mapping capability of BP neural network algorithm.This paper analyzes characteristics of customer,reduces the customer characteristics by using rough set theory,divides the equivalence relations,establishes the initial topological structure of the BP neural network and divides the customer groups by using K-means algorithm.Through introducing rough set theory,this paper improves BP Neural Network Algorithm to accelerate the rate of BP network convergence and escape from local minimum points,and finally,by the use of Rosetta software and Matlab programming,completes the mapping model of customer base characteristics and service project facing on segmented customer groups.
出处
《武汉理工大学学报（交通科学与工程版）》
2013年第4期758-762,共5页
Journal of Wuhan University of Technology（Transportation Science & Engineering）
基金
国家自然科学基金项目(批准号:71071122)
湖北省科技厅攻关计划项目(批准号:2010S0014)
武汉理工大学自主创新研究基金项目(批准号:2012-ZY-029)资助
关键词
服务映射
BP神经网络
客户服务策略
粗糙集
service mapping
BP neural network
customer service strategies
rough set
