Fraud Detection

Before credit-card-fraud-mlops became a full pipeline, this is where the modeling started: notebook-driven data cleaning, exploratory analysis, and comparing classifiers like logistic regression and random forest for spotting fraudulent transactions.

Why: The starting point before credit-card-fraud-mlops became a full pipeline —wanted to prove out the model itself first, before building an API and DVC/MLflow around it.

Stack: Python, Jupyter Notebook, scikit-learn