Guilherme Duarte Marmerola

My collection of causal inference algorithms built on top of accessible, simple, out-of-the-box ML methods, aimed at being explainable and useful in the business context

Random stuff I've been working on

Big Data's open seminars: An Interactive Introduction to Reinforcement Learning

Code for "Training models when data doesn't fit in memory" post

More about the exploration-exploitation tradeoff with harder bandits

Projeções de Investimentos no Tesouro Direto com R

My explorations regarding bayesian networks and structural causal models

Solução para o desafio "Santander Customer Satisfaction" no Kaggle

Code for treatment effect estimation

Improving XGBoost survival analysis with embeddings and debiased estimators

Trying out forest-based data representations

My webpage drawing inspiration from the Sleek theme: https://github.com/janczizikow/sleek

Simple app that predicts a dog's breed by searching the Stanford Dogs Dataset for similar dogs

Code for Article "Solving the mystery of my dog's breed with ML"

Configuração automática do t-SNE (Projeto de conclusão de curso Unicamp 2017)

Solução para o desafio "What's cooking?" no Kaggle

Solução para o desafio "Titanic: Machine Learning from Disaster" no Kaggle

Análise da remuneração dos funcionários da Unicamp (junho/2015)