Journal of Japan Industrial Management Association
Online ISSN : 2432-9983
Print ISSN : 0386-4812
Generation of a New Dispatching Rule for the Mean Tardiness Job Shop Scheduling Problem through the Application of Machine Learning Concepts
Dong-Ming HUANGTakao ENKAWAMasao AKIBA
Author information
JOURNAL FREE ACCESS

1991 Volume 41 Issue 6 Pages 383-389

Details
Abstract
There is no single dispatching rule which performs dominantly better than any other rules for the mean tardiness job shop scheduling problem since the perfomance of dispatching rules depends on the conditions of each problem. This paper suggests an approach to generate new robust dispatching rules.First, this approach learns from successful examples by applying customary dispatching rules to obtain good conflict resolution rules.Then, an inductive procedure is applied to generalize this set of accumulated rules.Applying this approach to the static job shop scheduling problem, a novel dispatching rule was generated and demonstrated to have better performance than other dispatching rules for a wide range of due date tightness.
Content from these authors
© 1991 Japan Industrial Management Association
Previous article Next article
feedback
Top