Started working on a Pet Project for Classification.
Few highlights from the start:
Few highlights from the start:
- EDA is important and cannot be dismissed as it will allow you to see what is going on with your data before building any model.
- When trying to discard or drop outliers, check if those that are supposed to be outliers belong to the positive class.
- Try to make your methods as modular as possible so that you can re-use them for further projects.
- Checkout if someone else has explored this data already and how they have done it.