Online learning for evolving demand

Model choice is a property of the data, horizon and loss function. This article uses online learning for evolving demand to highlight those design choices and their consequences.

Practitioner lens

For online learning for evolving demand, validation must respect the information set available at each forecast origin. Features published after the decision date create leakage and overstate likely performance.

First experiment

Report uncertainty, compute cost and maintenance burden alongside point accuracy.

This is an original educational brief. It is a starting point for discussion, not a claim that one technique fits every supply chain.