LightGBM regression on rank-transformed target: “No further splits with positive gain” and nearly-constant predictions
Summary This incident documents why a LightGBM regressor trained on a rank‑transformed target can collapse into “No further splits with positive gain” and produce near‑constant predictions. The failure is rooted in how LightGBM evaluates split gain, how rank‑scaled targets behave statistically, and how certain parameter combinations suppress variance to the point where the model cannot … Read more