Data Storm 1.0 - Business Problem - Team Fixzels
- Use case definition - A default can occur when a borrower is unable to make timely payments, misses payments, or avoids/stops making payments
- Business problem - As a result of credit card default, bank accounts close, non-performing loans and bad debts get written-off. Therefore, the bank needs help in predicting and preventing credit card default to improve their Bottom line.
- Key Objective - Finding out priority clients who have the highest risk of credit card default.
- credit_card_default_train.csv - the training set
- credit_card_default_test.csv - the test set
- Client_ID - customer id
- Balance_Limit_V1 - credit card limit given
- Gender - F = female, M = male
- EDUCATION_STATUS - education status of the customer
- MARITAL_STATUS - 1= single, 2 = married, 3 = others
- AGE - Age of customer
- PAY_JULY,PAY_AUG,PAY_SEP,PAY_OCT,PAY_NOV,PAY_DEC - "History of past payments" (-2 = paid two months in advance), (-1 = paid one month in advance), (0 = payment due current month), ( 1 = payment delay for one month), (2 = payment delay for two months), (8 = payment delay for eight months), (9 = payment delay for nine months and above)
- DUE_AMT_JULY, DUE_AMT_AUG, DUE_AMT_SEP, DUE_AMT_OCT, DUE_AMT_NOV, DUE_AMT_DEC - "Due amount of bill statements"
- PAID_AMT_JULY, PAID_AMT_AUG, PAID_AMT_SEP, PAID_AMT_OCT, PAID_AMT_NOV, PAID_AMT_DEC - "Amount paid in each month"
- NEXT_MONTH_DEFAULT - target variable - Predict if next month payment is a default payment or not
Data Storm 1.0 Competition Link : https://www.kaggle.com/c/data-storm-10/