Custom implementation of Naïve Bayes and Decision Tree classifiers from scratch, applied to categorical datasets (Mushroom & Congressional Voting). No scikit-learn used—focus on core ML principles like entropy, Laplace smoothing, and tree traversal.

🏛️ Interactive dashboard analyzing voting behavior and government alignment in the Brazilian Chamber of Deputies. Features nominal vote tracking, thematic classification, and high-performance data processing with R, Shiny, and Arrow.