Subscription Benefits
Case Study

CAN AN AUTOMATED MODEL BE BUILT FOR PREDICTING THE CHANCES OF SANCTIONING A LOAN?

Dr.Srilakshminarayana G, Associate Professor - Quantitative MethodsSDM Institute for Management Development, Mysore
Loan application, Bank, Legal and Non-Legal Issues,Automation, Machine Learning, Logistic regression, Prediction, Classification, Decision tree.

Abstract

Building a house is adream for every individual and looks forward to achieving their dream by working hard. Among other factors, financial aspects act as a hurdle to fulfill this dream. It is a regular practice for one to approach the banks or other financial institutions for a loan. Daily, hundreds of customers apply for the loan and they get scrutinized for deciding on them. While few applications get accepted, others get rejected.

Few applications that get rejected pass through the first stage of scrutiny and get rejected in the later stages. For example, rejection may be due to legal issues or lack of proper documentation. It is a difficult situation for both the banks and the customers to take the rejection. While customers look for receiving the loan amount, banks look for building a customer base. Scrutinizing the applications manually is time-consuming and customers expect a quick decision on their application so that, they can act appropriately and resubmit the application. Due to the delay in the response from the banks or the institutions, customers feel that they are losing their time. Similarly, Banks or other institutions feel that they must be extremely careful in giving a loan and follow a policy of screening the application in depth.

They also feel that the time taken to pass the application is acting as a hurdle in increasing their customer base. One of the solutions is to build an automated system that will help the customers to get their application status quickly andfor the banks to clear the application at the earliest. The current caselooks at providing the process that can be used to build anautomated model that scrutinizes the application at the first stage and sends a message to the customers on the status and help the banks to get the immediate status. Logistic regression is used for building a predictive model and a classification process, to get the scrutiny done in the first stage and use decision trees to find the status in the second stage.The case discusses the application of logistic regression and decision trees for building the automated model. It can be used by the faculty while discussing the applications of these decision-making methods.

Learning objectives

Increasing the customer base is an important target for financial institutions. Among the products offered, they attract customers by giving notifications regarding housing loans at lesser interest rates. Many customers tend to apply for the same and, sometimes many of the applications get rejected in the second or third stage of the loan process. This leads to customer displeasure and also loss of customers for the banks or the institutions. Hence, an automated model that scrutinizes the application in the initial stages and cautions the customers would help both the banks and the customers. This is the main issue in the current case.

This case study helps the faculty to explain the importance of Logistic regression in machine learning and classification. Also, the importance of decision trees in arriving atappropriate decisions. It helps the students to relate themselves with the situations they face in life concerningthe rejection of loans they apply for and understand the importance of having an automated model in decision making. It also helps them understand the application of statistical methods in decision-making. Hence, the learning objectives include

a. Understanding the importance of having an automated model while taking decisions on loan applications.
b. Learn the process of building a Predictive model for arriving at decisions on the applications in the initial stages and a classification tree in the later stages.
c. Learn how Logistic regression can be used in classification problems. Also, how it can be used in machine learning for building the automated model.
d. Discuss how the model built, gives accurate results for taking a final decision on the applications received.
e. Learn how to integrate the results with other disciplines such as HR and Marketing.

  • Pub Date:
    25 Aug 2021
  • Source:
    SDMIMD
  • Discipline:
    Marketing Management,Human Resource Management,Financial Management
  • Product#:
    1635
  • Keywords:
    Loan application, Bank, Legal and Non-Legal Issues,Automation, Machine Learning, Logistic regression, Prediction, Classification, Decision tree.
  • Length:
    Pdf : 21 page(s) ,

Related Items

x