Non-cost-based lot-sizing heuristics: A comment on Cheng

Abstract

In a recent article, Cheng [1] argued that previous economic orde quantity (EOQ) related heuristics are inherently flawed, and proposed two lot- sizing heuristics which are non-cost-based (cost data ignored in determining lot-sizes) and relatively free from EOQ considerations. Based on a computational study, it was concluded that non-cost-based heuristics“can be quite effective for lot-sizing decisons” [1, p.95] as compared to optimal solutions generated by the Wagner-Whitin algorithm [2]. This comment critiques the premise under which the Cheng study was conducted, illustrates the inadequacy of the test data and errors in the computational results, and compares the performance of the proposed heuristics to two EOQ-based heuristics. It is concluded that the non-cost-based heuristics proposed in Cheng [1] are naive and arbitary in nature, and that their computational performance is far inferior to existing heuristics in the literature.