The dataset can be found on kaggle (https://www.kaggle.com/datasets/andrewmvd/divorce-prediction)
The data processing, model creation and model training finds place in the divorce.ipynb. After training i did some feature importance calculations and indeed some features doesnt seem to have an impact, so they could be removed.
To start the webapp created with gradio:
python gradio_app.py
Now you can visit the URL presented by you in the terminal and try out the divorce predictor.
