Kontaktujte nás | Jazyk: čeština English
Název: | Insight into adaptive differential evolution variants with unconventional randomization schemes | ||||||||||
Autor: | Šenkeřík, Roman; Viktorin, Adam; Kadavý, Tomáš; Pluháček, Michal; Zelinka, Ivan | ||||||||||
Typ dokumentu: | Článek ve sborníku (English) | ||||||||||
Zdrojový dok.: | Communications in Computer and Information Science. 2020, vol. 1092 CCIS, p. 177-188 | ||||||||||
ISSN: | 1865-0929 (Sherpa/RoMEO, JCR) | ||||||||||
Journal Impact
This chart shows the development of journal-level impact metrics in time
|
|||||||||||
ISBN: | 978-3-03-037837-0 | ||||||||||
DOI: | https://doi.org/10.1007/978-3-030-37838-7_16 | ||||||||||
Abstrakt: | The focus of this work is the deeper insight into arising serious research questions connected with the growing popularity of combining metaheuristic algorithms and chaotic sequences showing quasi-periodic patterns. This paper reports an analysis of population dynamics by linking three elements like distribution of the results, population diversity, and differences between strategies of Differential Evolution (DE). Experiments utilize two frequently studied self-adaptive DE versions, which are simpler jDE and SHADE, further an original DE variant for comparisons, and totally ten chaos-driven quasi-random schemes for the indices selection in the DE. All important performance characteristics and population diversity are recorded and analyzed for the CEC 2015 benchmark set in 30D. © Springer Nature Switzerland AG 2020. | ||||||||||
Plný text: | https://link.springer.com/chapter/10.1007/978-3-030-37838-7_16 | ||||||||||
Zobrazit celý záznam |