Back

ML classification / Complete

MentalHealth CatBoost pipeline

CatBoost pipeline on self-reported questionnaire data with logistic baseline, SHAP and equity analysis.

CatBoostOptunaSHAPFairness

Problem

Handle a sensitive categorical classification task without turning it into a medical claim.

Approach

Native categorical CatBoost, stratified 5-fold validation, Optuna logloss, JSON/Markdown exports and gender/country analysis.

Evidence

Complete CLI pipeline.

GitHub Actions install-check.

Interpretability and bias analysis included in scope.

Limitations

ML modeling project, not a clinical tool or medical recommendation.

More work

Back