What is quantile regression in ensemble models?

Quantile regression estimates conditional quantiles instead of mean predictions.

Quantile regression in ensembles allows prediction of uncertainty intervals rather than single point estimates. For example, predicting the 10th, 50th, and 90th percentile gives a range of possible outcomes. Gradient boosting frameworks like LightGBM and XGBoost support quantile loss functions. This is useful for risk-aware decision-making where uncertainty matters.