How the Landscape of Random Job Shop Scheduling Instances Depends on the Ratio of Jobs to Machines - Robotics Institute Carnegie Mellon University

How the Landscape of Random Job Shop Scheduling Instances Depends on the Ratio of Jobs to Machines

M. J. Streeter and S. F. Smith
Journal Article, Journal of Artificial Intelligence Research, Vol. 26, No. 1, pp. 247 - 287, May, 2006

Abstract

We characterize the search landscape of random instances of the job shop scheduling problem (JSP). Specifically, we investigate how the expected values of (1) backbone size, (2) distance between near-optimal schedules, and (3) makespan of random schedules vary as a function of the job to machine ratio (N/M). For the limiting cases N/M approaches 0 and N/M approaches infinity we provide analytical results, while for intermediate values of N/M we perform experiments. We prove that as N/M approaches 0, backbone size approaches 100%, while as N/M approaches infinity the backbone vanishes. In the process we show that as N/M approaches 0 (resp. N/M approaches infinity), simple priority rules almost surely generate an optimal schedule, providing theoretical evidence of an "easy-hard-easy" pattern of typical-case instance difficulty in job shop scheduling. We also draw connections between our theoretical results and the "big valley" picture of JSP landscapes.

BibTeX

@article{Streeter-2006-120456,
author = {M. J. Streeter and S. F. Smith},
title = {How the Landscape of Random Job Shop Scheduling Instances Depends on the Ratio of Jobs to Machines},
journal = {Journal of Artificial Intelligence Research},
year = {2006},
month = {May},
volume = {26},
number = {1},
pages = {247 - 287},
}