|Appears in Collections:||Computing Science and Mathematics Conference Papers and Proceedings|
Kwan, Raymond S K
|Title:||A fuzzy evolutionary approach with Taguchi parameter setting for the set covering problem|
|Citation:||Li J & Kwan RSK (2002) A fuzzy evolutionary approach with Taguchi parameter setting for the set covering problem. In: Proceedings of the 2002 Congress on Evolutionary Computation. CEC'02. 2002 World Congress on Computational Intelligence - WCCI'02, Honolulu, HI, USA, 12.05.2002-17.05.2002. Piscataway, NJ, USA: IEEE. https://doi.org/10.1109/cec.2002.1004414|
|Conference Name:||2002 World Congress on Computational Intelligence - WCCI'02|
|Conference Dates:||2002-05-12 - 2002-05-17|
|Conference Location:||Honolulu, HI, USA|
|Abstract:||The set covering problem has a very wide area of applications, and scheduling is one of its most important applications. At CEC2001, a fuzzy simulated evolution algorithm was reported for a set covering problem in bus and rail driver scheduling. This paper reports on the generalisation of the approach to the class of set covering, which is basically to cover the rows of a zero-one matrix with a subset of columns at minimal cost. The simulated evolution method iteratively transforms a single solution treating its selected columns as a population. At each iteration a portion of the population is probabilistic ally, biased towards selecting the 'weaker' columns, discarded; and the broken solution is then reconstructed by means of a greedy heuristic. The columns are evaluated under several criteria formulated based on fuzzy set theory. The set of weights associated with the evaluation criteria were previously either fixed in advance or calibrated by means of a simplified genetic algorithm. In this paper, we consider some discrete levels of values instead of a continuous range from 0 to 1 for each weight. We have also included two other parameters driving the evolutionary process as additional factors having some discrete levels of values. Taguchi's orthogonal experimental design is applied. This has the effect of comprehensively evaluating the combinations of factors, although only a small fraction of the possible combinations is explicitly experimented. Better results have been obtained, and comparisons with those previously reported are discussed.|
|Status:||VoR - Version of Record|
|Rights:||The publisher does not allow this work to be made publicly available in this Repository. Please use the Request a Copy feature at the foot of the Repository record to request a copy directly from the author. You can only request a copy if you wish to use this work for your own research or private study.|
|01004414.pdf||Fulltext - Published Version||637.67 kB||Adobe PDF||Under Permanent Embargo Request a copy|
Note: If any of the files in this item are currently embargoed, you can request a copy directly from the author by clicking the padlock icon above. However, this facility is dependent on the depositor still being contactable at their original email address.
This item is protected by original copyright
Items in the Repository are protected by copyright, with all rights reserved, unless otherwise indicated.
If you believe that any material held in STORRE infringes copyright, please contact email@example.com providing details and we will remove the Work from public display in STORRE and investigate your claim.