Machine learning and data science projects from coursework and independent research.
Converted thousands of inconsistent PDF complaints to CSV format and trained an XGBoost model to predict complaint outcomes using NLP features. Applied SHAP interpretability to understand model predictions.
Formatted a large public data set of court case descriptions and their topics using Pandas. Trained models with Multinomial and Gaussian Naive Bayes and Logistic Regression. Tested model accuracy across classifier type and data-cleaning techniques.