This is an End-to-End Machine Learning project with deployment. The project aims at developing a model and predicting the Credit card Defaulter based on his past 6 months transaction information.
The project was created using Python Flask and deployed in Azure App services.
Link: https://credit-card-default-prediction.azurewebsites.net
As we are seeing with the growing number of credit card users, banks & credit card companies are facing an increase in credit card default rate. This increase in defaults creates much impact to the financial companies along with the cardholders and pulls them to losses. As such, data analytics can provide solutions to manage the current phenomenon and management of credit defaults. This project discusses the implementation of a model which predicts if a given credit card holder has a probability of defaulting in the following month, using their demographic data like (age, gender, marital status) and behavioural data from the past 6 months.
In the dataset we have 25 columns which reflect various attributes of the customer. The target column is default.payment.next.month which tells whether a person will default or not. In this dataset, we have information on default payments, demographic factors, credit data, history of payment, and bill statements of credit card holders from April 2005 to September 2005.
Datasource Link: https://www.kaggle.com/datasets/uciml/default-of-credit-card-clients-dataset
Demo Video Link: https://youtu.be/qn8MmiEdNao
- numpy
- pandas
- matplotlib
- sklearn
- Flask
- gunicorn
- imblearn
- xgboost
- mysql-connector-python
- six
- dill
- Creating conda environment
conda create -p venv python==3.7 -y
- For activating environment
"conda activate venv/"
- For installing packages in requirements.txt
pip install -r requirements.txt
- To add files to git
git add <file_name> => for adding single file to git.
git add . => for adding all the files to your local git.
- To check the git status
git status
- To create version/commit all the changes to git
git commit -m "Message"
- To send version/changes to github.
git push origin main
-
EDA&Feature_engineering.ipynb
In this Notebook, I have performed complete End-to-End Machine Learning process from DataIngestion to Data Evaluation
- Steps Performed:
- Data Ingestion
- EDA
- Feature Engineering
- Feature Selection
- Sampling
- Model Training
- Best model using(AUC-ROC) curve
- Steps Performed:
2. CreditCard_DefaultPrediction.ipynb
This is the main syntax for the project i.e, main steps to be implemented in the project.
@Ravi Teja Mandarapu