What ensemble methods did you use for demand forecasting?medium
Answer
Used an ensemble of XGBoost, a statistical baseline (exponential smoothing or ARIMA), and a linear model for regularization.
Explanation
Used an ensemble of XGBoost, a statistical baseline (exponential smoothing or ARIMA), and a linear model for regularization. Combined via weighted averaging or stacking where a meta-learner decides weights based on recent forecast error. Reduces the risk of any single model failing on edge SKUs.
Follow-upCan you give a production example?