A fuzzy support vector machine algorithm and its application in telemarketing

Abstract

Telemarketing applications in various industries are increasingly popular in modern society. Telemarketing is one of the important means to contact with customers in insurance companies, banks and other financial systems. For example, if we accurately predict telemarketing successfully, we can appropriately reduce the cost and the scope of marketing in banking, which has very important significance. In this paper, the authors present a fuzzy support vector machine (SVM) algorithm, based on the data of customer information obtained in telemarketing campaigns in a financial institution of Portugal, using Weka and Matlab software, predict the success of telemarketing. Experimental results show that the fuzzy SVM algorithm outperforms the traditional SVM with 92.89% predicting accuracy rate.