An adaptive approach for solving dynamic scheduling with time-varying number of tasks - Part II
Date
2011
Authors
Abello, M.
Bui, L.
Michalewicz, Z.
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Conference paper
Citation
Proceedings of the 2011 IEEE Congress of Evolutionary Computation (CEC 2011), held in New Orleans, LA, USA, 5-8 June, 2011: pp.1711-1718
Statement of Responsibility
Manuel Blanco Abello, Lam Thu Bui and Zbignew Michalewicz
Conference Name
IEEE Congress of Evolutionary Computation (2011 : New Orleans, USA)
Abstract
Changes in environment are common in daily activities and can introduce new problems. To be adaptive to these changes, new solutions are to be found every time change occur. This two-part paper employs a technique called Centroid Based Adaptation (CBA) which utilize centroid of non-dominated solutions found through Multi-objective Optimization with Evolutionary Algorithm (MOEA) from previous environmental change. This centroid will become part of MOEA's initial population to find the solutions for the current change. The first part of our paper deals mainly on the extension of CBA, called Mapping Task IDs for CBA (McBA), to solve problems resulting from time-varying number of tasks. This second part will show the versatility of McBA over a portfolio of algorithms with respect to the degree of changes in environment. This demonstration was accomplished by finding a model relating the degree of changes to the performance of McBA using Nonlinear Principal Component Analysis. From this model, the degree of change at which McBA's performance becomes unacceptable can be found. Results showed that McBA, and its variant called Random McBA, can withstand larger environmental changes than those of other algorithms in the portfolio.
School/Discipline
Dissertation Note
Provenance
Description
Access Status
Rights
©2011 IEEE