Polynomial time algorithms for minimax regret robust uncapacitated lot sizing models

D. Li, D. Romero Morales

We study the Minimax Regret Uncapacitated Lot Sizing (MRULS) model, where the production cost function and the demand are subject to uncertainty. We propose a polynomial time algorithm which solves the MRULS model in O(n6) time. We improve this running time to O(n5) when only the demand is uncertain, and to O(n4) when only the production cost function is uncertain.

Key words: robust optimization, minimax regret, lot sizing, production cost and demand uncertainties

go to main page