Multi-objective Dynamic Analysis Using Fractional Entropy

• Published in: Intelligent Systems Design and Applications Vol. 557; pp. 448 - 456
• Main Authors: Solteiro Pires, E. J., Tenreiro Machado, J. A., de Moura Oliveira, P. B.
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2017; Springer International Publishing
• Series: Advances in Intelligent Systems and Computing
• Subjects: Artificial intelligence; Convergence; Diversity; Dynamic; Fractional entropy; Multi-objetive genetic algorithm
• ISBN: 9783319534794; 3319534793
• ISSN: 2194-5357; 2194-5365
• DOI: 10.1007/978-3-319-53480-0_44

Multi-objective optimization evolutionary techniques provide solutions for a specific problem using optimally concepts taking into consideration all the design criteria. In the last years, several multi-objective algorithms were proposed but usually the performance is measured at the end neglecting, therefore, the solution diversity along the interactions. In order to understand the evolution of the solutions this work studies the dynamic of the successive iterations. The analysis adopts the fractional entropy for measuring the statistical behavior of the population. The results show that the entropy is a good tool to monitor and capture phenomena such as the diversity and convergence during the algorithm execution.