Research Project Assistant — Machine Learning for Insurance Risk Prediction (Cathay Life Collaboration)
Worked as a Research Project Assistant building and evaluating machine learning models for an imbalanced insurance risk prediction task in collaboration with Cathay Life. The work included preparing large-scale health datasets and applying techniques to address class imbalance before training. Model performance was assessed using standard classification metrics to ensure reliable predictions. • Applied resampling methods (SMOTE, ENN) to improve class balance • Performed data cleaning, feature engineering, and feature selection • Trained and compared Random Forest and Gradient Boosting models • Evaluated results using AUC-ROC and F1-score
2019 - 2021