Abstract:   Covid-19 is an unexpected event that occurred in the beginning of 2020 at a place in the world and has spread all over the world very fast. The most suffered are the small and medium businesses. This has given opportunities for few financial firms to support them to come out of this situation and restart their businesses. Also, help few of them to recover and redefine their business goals. The challenge the financial firms face is the authenticity of the applications received from the businesses for loan and finding the chances of a customer becoming a defaulter.

Taking into consideration the recent developments in the technology and methods, and availability of the same for adoption, fintech companies are using the same in their decision making process. One of the key decisions, in the recent times, is on the applications they receive for the financial help. The financial organizations who are ready to lend have to check every aspect of the applicant with caution and then take a decision. Doing the same using the usual manual process, when there are hundreds of applications, is not a right choice. Hence, they need to take the advantage of the technology and the methods to classify the applications into two sets. The first set of applications are those whose chances of default will be low and the second set are those whose chances will be high that are not eligible (which includes modifications, rejections etc.). Another challenge they face is to have an automated process that can be used continuously, so that they can take quick decisions. Taking these two into consideration, the current case has been developed.

ATM Iinc. is one such financial organization that lends to the businesses, facing the challenge of classifying the applications. They have a very good reputation in the market and will be quick in their decisions in adopting the technology. They do not want to land into trouble due to a customer becoming a defaulter. They wish to establish a process that will help them to predict the chances of a customer becoming a defaulter and also aspects that have to be significantly considered while taking a decision on the applicants. The case revolves around this issue of classification and how a classification method, Chi-square Automatic Interaction Detection (CHAID), can be used to resolve the issue.

 Learning Objectives

1. Understanding the importance of having an automated model while taking decisions on customers.
2. Learn the process of building a Predictive model for calculating the chances.
3. Learn how CHAID can be used in classification problems and the difference between this method and other methods.
4. Understand the differences in the strategies adopted by the organizations that use predictive models and those that do not use predictive models.