Applying AI Techniques to Predict the Success of Bank Telemarketing

Abstract

Telemarketing is one of the marketing methods valued by enterprises in marketing tools. Compared with general channel sales, it not only has lower cost, but also has better interaction with customers, and the entire sales process is faster. And where necessary processing huge owned material times, how to improve at the processing efficiency rate and the exploration of useful information is competitive advantage to the source, using AI technology to assist market segmentation and promotion is currently the commonly used method. Telemarketing is the use of information technology to carry out planned sales activities for the target market, helping the company maintain close ties with existing or potential customers, and increase sales and services. In this research, we use the data set of the direct marketing activities of Portuguese banking institutions in the UCI data bank as the data sample. We use python to build an AI prediction model, including Logistic Regression (LR), and a decision tree (DT) and support vector machine (SVM), three modeling results show that the accuracy of the way were 0.908 (LR), 0.887 (DT), 0.899 (SVM), with Logistic regression analysis performed the best.