Here is our team I_over_B solution. Please upvote the notebook!

overview:

1. We handled the imbalanced data by dropping unnecessary rows in both the train and test sets (the logic behind this is detailed in the notebook).
2. We used two different models: one for Kenya and one for Ghana.
3. We incorporated economic indicators for the Ghana model.
4. We did not apply any post-processing or use tricks like the "Type 3" thing..

Notebook:https://www.kaggle.com/code/zribiahmedyassine/i-over-b-updated1-african-credit-scoring-challenge