Simple streamlit application to interact with a ML classification model based on PKDD'99 default financial data.

To create the development environment it's recommended to use conda.

Run the following commands to get the environment ready

conda create -n ENVIRONMENT_NAME python=3.8
conda activate ENVIRONMENT_NAME
pip install -r requirements.txt

When running for the first time, dataset should be created and model should be trained with the following commands

python data/build_dataset.py
python model/train.py

To run the streamlit application in your local host use

streamlit run app.py

Production: https://default-prediction-app.herokuapp.com/