Predicting Efficiency of Direct Marketing Campaigns for Financial Institutions

Abstract

All marketing campaigns are dependent on the data that their clients provide. These datasets include everything from their name, number, their salary, the loans they’ve already taken, the money they have in their account etc. These datasets are huge and it is impossible for a human to analyze the patterns in the client deposits and whether the campaign will be a success or not. This paper introduces prediction of the success rates of the campaigns with the help of various machine learning predictive algorithms: Random forest, KNN and KNN using tensor flow framework. In the recent years, the success of general campaigns led by financial institutions have been declining and to step up their game and increase campaign effectiveness they need to target the clients very specifically. The predictive results obtained, with the highest accuracy, help in increasing the target audience and ensure that it will be a success for the financial institutions in picking out clients who will subscribe to the different marketing schemes. This paper uses multiple accuracy parameters to predict the potential success a direct marketing campaign will have on specific clients.