I made exploratory data analysis and modeling from CreditCustomer dataset, click here to preview the result. The best classifier in that analysis is Random Forest Classifier.
In this repositories, I uploaded the flask_app.py (open github code or open in pythonanywhere) and model_randomforest.pkl (open github code)files
Now, you can use it for your company in the easiest way. First of all, Install postman to your computer. You can download it here. Then, do the following instructions:
- You have to add (+),
- Change 'GET' to 'POST',
- Fill the blank in 3 dwilarasathina.pythonanywhere.com/api
- Choose 'Body'
- Choose 'raw'
- The last step is you have to input your data in box number (6).
In detail, look at the picture below 
Data you can input in box number (6) is customers data which include
- PAY_1 : Non negative, integer number. Payment in the 1st month by customers. 0 : Customer doesn't late to pay. 1 : Customer late 1 month to pay. 2 : Customer late 2 months to pay. and so on.
- PAY_2 : Non negative, integer number. Payment in the 2nd month by customers. 0 : Customer doesn't late to pay. 1 : Customer late 1 month to pay. 2 : Customer late 2 months to pay. and so on.
- LIMIT_BAL : Non negative number. Maximum Credit Limit.
Example. There are 3 customers, with data record as the following table.
| Customer ID | PAY_1 | PAY_2 | LIMIT_BAL |
|---|---|---|---|
| 1 | 4 | 2 | 200000 |
| 2 | 0 | 1 | 75000 |
| 3 | 2 | 1 | 50000 |
In the box number (6), please use this template.
[{"PAY_1":4,"PAY_2":4,"LIMIT_BAL":200000}, {"PAY_1":0,"PAY_2":1,"LIMIT_BAL":75000}, {"PAY_1":2,"PAY_2":1,"LIMIT_BAL":50000} ]
Then, you can click "Send" button, and you will get the result like this. 
From that result, we know classification from each customer. The result should be :
| Customer ID | PAY_1 | PAY_2 | LIMIT_BAL | RESULT |
|---|---|---|---|---|
| 1 | 4 | 2 | 200000 | Not Approved, Credit must be stopped |
| 2 | 0 | 1 | 75000 | Approved |
| 3 | 2 | 1 | 50000 | Not Approved, Credit must be stopped |