Predicting Customer’s Subscription Response to Bank Telemarketing Campaign Based on Machine learning Algorithms

The cut-throat competition in the banking industry has led to the paradigm shift of banking strategies from conventional methods towards state-of-the-art data mining and machine learning techniques. Over the years, the banking industry has shifted from personal individual visits towards telemarketing which is an effective and cheaper way of carrying out the marketing campaigns through telecommunication. However, the banks are always seeking new methods and techniques to help them advertise the selected products to the interested customers in order to minimize the time and cost spent on marketing with maximum number of customers subscribing to the new products, consequently increasing their profit value. This paper attempts to provide the banks with a data mining solution to predict the customer’s response to the telemarketing campaign using an ensemble classifier based on hybrid machine learning models. A thorough Exploratory Data Analysis (EDA) is also performed to give useful insights about the customer’s attributes leading to a higher subscription rate. Our ensemble classifier caters the class imbalance problem in the dataset and predicts the customer’s response with an accuracy of 95.6%.