Telemarketing Success: Evaluation of Supervised Classifiers

Nowadays telemarketing constitutes a way in which goods and services companies can access to possible potential customers through phone calls. Telemarketing campaigns are focused on offer to potential customers or users, contracting or buying a good or service. Ascertain a priori which phone calls will be successful is a competitive advantage to the companies due to this allow them to reduce costs and focus on most likely groups of potential customers which would contract or buy the goods or services offered. For this task it is necessary to classify the phone calls in successful and unsuccessful calls, which is possible using supervised classifier. In this paper, we tested some supervised classification algorithms and compare their performance, based on the Area under the ROC Curve, over different well-known telemarketing datasets.