This repository contains credit scoring prediction model using random forest algorithm. The model is deployed by using Flask framework. You can run this model by using postman application.
This repository has two folders :
1. code
- Credit-Score.py : model python.
- flask_app.py : server python to be uploaded on Flask.
- random_forest.pkl : random forest algorithm in pickle.
- request.py : request model in python.
2. dataset
- training.csv : dataset to build the model.
The steps and requirements needed to run the credit scoring prediction will be explained on the next section.
To use this credit scoring prediction model, you should have :
- Postman install here.
1. Open the postman, then set the request method to POST.
2. Input the API link hmaghfira.pythonanywhere.com/api.
3. Open the Body tab below the API link column, click raw and choose JSON as type of the input request.
4. Fill the body with the request script. The example of the script is here. The guidance for input parameters, go to this section
5. After that, click SEND button beside of API link column (the button is blue) to run the script.
6. On the below of the body script, you will see the result of the credit scoring prediction.
| NAME | EXPLAIN | EXAMPLE OF INPUT |
|---|---|---|
| LIMIT_BAL | The limit of the credit balance | The amount of rupiah : 30000 (means Rp. 30.000) |
| PAY_1 | The status whether the customer pay on time or not | 0 : On time, 1 : late |
| AGE | The age of the customer | Integer : 22 (means 22 years old) |
| EDUCATION | The education level of the customer | 1 : S2/S3, 2 : S1, 3 : SMA, 4 : others |
| SEX | The gender of the customer | 1 : Man, 2 : Woman |
Hania Maghfira, Astra Data Scientist Bootcamp 2019.
hania.maghfira@ai.astra.co.id