Gary Vidal

Bayesian Classification Trees

Implementation of Bayesian classification models with CART trees and Markov-chain-based optimization to improve dataset alignment.

Minimal classifier tree diagram

Project Overview

This project explores a Bayesian formulation for CART classification trees, aiming for stronger fit and generalization compared with traditional deterministic approaches.

It studies optimization strategies using Markov chains and dataset-alignment criteria to analyze model behavior across different classification scenarios.

Tech Stack

Python
scikit-learn
Bayesian Methods
Markov Chains