/ds-ml-project

Primary LanguageJupyter Notebook

ds-ml-project

For our first ML project, our group worked on a Zindi challenge for Credit Fraud. The stakeholders requirement was to maximise the F1-Score metrics.

Challenges

  • dealing with highly imbalanced data (only 0.2% of all cases were fraudulent)
  • we used SMOTE algorithm for re-scaling
  • create additional insights through feature engineering
  • especially, handling date-time object
  • work out behavioral patterns from the given data

Models tested and employed

  • dummy classifier to guess minority class (maximises F1-Score) as a base line model
  • decision tree
  • random forest
  • AdaBoost
  • stacking (decision tree, random forest, AdaBoost, meta: logistic regression)

Environment

pyenv local 3.9.4
python -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install -r requirements.txt