The Predictive Analytics Models Development to Predict a Telemarketing Activities Success for the Commercial Bank

In the framework of this work, a study of telemarketing activities is conducted, and an approach is proposed to optimize and improve the efficiency of the telemarketing department of a commercial bank using artificial intelligence and modern machine learning algorithms. The goal of the work is to development several computer-based machine learning predictive models for solving the problem of predicting the success of the telemarketing activity of the bank. The subject of the study is telemarketing activity in the banking sector. The subject of the research is the process of telemarketing management in financial structures. In the course of the work the following tasks were accomplished: 1. The process of telemarketing as a whole has been studied; 2. Features of telemarketing activity in the banking sphere were studied; 3. The analysis of the available data set on telemarketing activity of the bank was carried out; 4. Predictive models for predicting the success of telemarketing activity have been developed; 5. A comparative analysis of models on a test sample of data was conducted to determine the best; Analysis of the available data on the marketing campaigns of the bank made it possible to determine that the outcome of the call is affected not only by personal information about clients or information about previous communications during the current marketing campaign. The client's inclination to open a deposit can be influenced by various external factors that show the overall economic and social conditions of the region in which the client lives. The use of these factors in machine learning algorithms allowed the construction of several predictive models capable of predicting call outcomes with good accuracy. In the future, the developed models can be integrated into CRM or ERP systems as separate prognostic modules.