CatBoost vs. XGBoost vs. LightGBM on UCI Credit Card Default data, optimized for recall on the minority class. SHAP explainability and EU AI Act compliance documentation built in.

Most marketing analytics measures who converted, not who converted because of the call. Built on the UCI Bank Marketing dataset with covariate adjustment, segmentation, and SHAP explainability.

How much genetic architecture is shared across ADHD, MDD, ASD, and bipolar disorder, and how have those estimates shifted with the 2023 iPSYCH / PGC ADHD GWAS? Personal interest project.

LA's shelter system is hugely overcrowded. Which intake characteristics most predict whether an animal gets adopted, transferred, or euthanized, and how long they wait for an outcome? Public data, public-facing visualizations.

MLB's pre-arbitration service time rules create a natural experiment. Which pre-arb performance signals causally predict post-arb value, and which signals do front offices act on that turn out not to matter?

Which biomechanical and performance metrics decline first, before the headline output stats start moving, and the market prices it in? Using causal structure learning on Statcast data to find leading indicators aging curves miss.

A softball coach told me my whole life not to swing on a 2-0 count. Capstone thesis project using Markov Chains on MLB data to test whether he was right. He was. This is also how I ended up in data science.