/migros-datathon

Primary LanguageJupyter Notebook

preparation

kaggle competitions download -c migros-datathon-coderspace
python -m venv .venv
.venv/bin/activate
pip install -r requirements.txt

First run preprocess_transaction_sale.py to generate more useful version of the transaction_sale.csv and then preprocess.py to generate the final dataframes.

plan

Deal with imbalanced data

oversampling and undersampling data augmentation

Feature engineering

Ensemble

VotingClassifier