UTS Pemrograman Sains Data

End‑to‑end comparison of Random Forest and Gradient Boosting classifiers on an imbalanced phishing‑email dataset. We split 60/20/20, applied down‑sampling, up‑sampling, ROSE and SMOTE, tuned probability thresholds to maximize F₁, then evaluated the champion model on a held‑out test set

Análise de aceitabilidade da desigualdade a partir do Latinobarômetro - Rodada 2023

DS2003 Final Project

Problems 3, 8, and 9