Our project for the UN Datathon 2023 explores using regional data to predict GDP. We use data for all of the counties of Kenya to predict GDP values. We then use LIME to interpret the model outputs and provide insights into investments that will most benefit a region.
- This project is written in Python, and a requirements.txt file is provided.
- The notebooks to create the model can be found in the model_notebooks repo.
- The data the model uses is in the datasets or EDA folders.
- The model outputs including pickle files of the model and min-max scalar are in the model_outputs folder.
- An app has been developed. The code for this can be found in the data_app folder and also in a separate repo.
- The app is also available in a separate repo and deployed to be explored here