Back

ML benchmark / Complete

HumanWeight regression benchmark

Ridge, RandomForest and LightGBM benchmark with preprocessing, Optuna, metrics and interpretability.

LightGBMRandomForestSHAPOptuna

Problem

Build a reproducible and interpretable ML pipeline on a tabular regression task.

Approach

EDA, cleaning, splits, Optuna tuning, training, SHAP, permutation importance and linear baseline.

Evidence

RandomForest test R² 0.9977 in the README snapshot.

LightGBM test R² 0.9967.

Explicit comparison with a Ridge baseline.

More work

Back