The purpose of the case study is to identify the candidates who have the highest likelihood of failing on their loan repayment.
The project is to look at a loan dataset for a consumer lending company that needs to decide if it should give a loan to a person or not. As part of the case study, we are looking at applicants whose loans have been charged off. We do this by looking at their annual income, loan amount, term, and other factors.
- The business problem we're trying to solve is how to predict or better yet, lower the loss of money caused by lending money to someone who isn't likely to pay it back. The dataset is a loan dataset that includes information like loan term, house ownership, proof status, and more.
• We need to be cautious while approving higher loan amounts, particularly for borrowers with higher risk profiles (e.g., low credit grades, high DTI, low income etc.). • As higher grades are less likely to be defaulted, prefer customer with grade A,B. • Prefer customers who are already home-owners. • Customers with higher & stable income can be prioritized. • Loans with 36 months term have more chances to be paid back compared to 60 months. • Loans where applicant’s source of income is verified are more reliable. • Loan applicants with history of bankruptcy are less preferable. • Loans with interest rates between 7.5% to 11% are more preferable.
- Python Version 3.11.x
- Pandas
- Matplotlib
- Seaborn
We would like to thanks our Professors at IIIT-B and Upgrad. Without them this case study will not be possible.
Created by : Anand Maitrey Satyam Singh For Upgrad ML C66 Batch