- Developing a basic understanding of risk analytics in banking and financial services and understand how data is used to minimize the risk of losing money while lending to customers.
- Examining the impact of variables such as loan type, loan purpose, business or commercial nature, and credit score on loan defaults.
- Investigating the correlation between upfront charges, loan amount, interest rates, and property values with the likelihood of default. -
- Analyzing patterns and uncovering insights into default tendencies.
| Column Name | Description |
|---|---|
| ID | Unique identifier for each row |
| year | Year when the loan was taken |
| loan_limit | Indicates if the loan limit is fixed or variable: cf - confirm/fixed, ncf - not confirm/not fixed |
| Gender | Gender of the applicant: male, female, not specified, joint (in case of applying as a couple) |
| loan_type | Type of loan (masked data): type-1, type-2, type-3 |
| loan_purpose | Purpose of the loan (masked data): p1, p2, p3, p4 |
| business_or_commercial | Specifies if the loan is for a commercial establishment or personal establishment |
| loan_amount | Amount of the loan |
| rate_of_interest | Interest rate applied to the loan |
| Upfront_charges | Down payment made by the applicant |
| property_value | Value of the property for which the loan is taken |
| occupancy_type | Occupancy type for the establishment |
| income | Income of the applicant |
| credit_type | Credit type of the applicant: EXP, EQUI, CRIF, CIB |
| Credit_Score | Credit score of the applicant |
| co-applicant_credit_type | Credit type of the co-applicant |
| age | Age of the applicant |
| LTV | Loan-to-value ratio of the applicant |
| Region | Region of the applicant |
| Status | Loan status: 1 - defaulter, 0 - normal |
- Data loading and exploaration
- Data cleaning
- Feature Enginnering
- Univariate Analysis
- Bivariate Analysis
- Multivariae Analysis
- Impact of ddifferent variabes on defaulters
- Insights
- Key Findings
- Recommendations
- You can access the full Python analysis on Google Colab using the following link: View the notebook
A detailed analysis report is available in the following PDF file: View Report.