Deploy Machine Learning Model for Effective Bank Telemarketing Campaign

Abstract

Marketing is a strategy that brings goods and services to all kinds of people. Most of the time marketing is not done to target the audience but to everyone to increase the chance of people opting for that particular product or service. Organizations use various marketing techniques like social media, telemarketing, advertisements, etc. The banking industry frequently chooses to speak with the user directly. The majority of the time, the customer hangs up as soon as they realize it was a telemarketing call because they are irritated. This is a result of a lack of essential client information. A deep learning model that analyses numerous parameters and forecasts whether a person is in a position to choose a loan or deposit money in the bank was developed to address this issue. For this purpose, a database of various parameters that the individual's demands depend on is compiled from Kaggle. On this dataset, several preprocessing approaches are then used. The procedures involve the elimination of missing values, category characteristics, outliers, undesirable features, and data balancing. To produce three machine learning (ML) models, three different techniques were applied. For this, the linear regression algorithm, the K-nearest neighbor algorithm, and the random forest (RF) algorithm are employed. To find the best algorithm for calling-list filtering, this is done. The results of testing these ML models using the preprocessed dataset are next analyzed. According to the results of the analysis, the RF algorithm is the best among the three algorithms. The RF algorithm's accuracy is somewhat more than 95%, which is higher than many other algorithms currently in use. In the future, the algorithm can be used as the backend process for a website or an application.