作者: Kamal Z. Zamli , Fakhrud Din , Bestoun S. Ahmed , Miroslav Bures
DOI: 10.1371/JOURNAL.PONE.0195675
关键词:
摘要: The sine-cosine algorithm (SCA) is a new population-based meta-heuristic algorithm. In addition to exploiting sine and cosine functions perform local global searches (hence the name sine-cosine), SCA introduces several random adaptive parameters facilitate search process. Although it shows promising results, process of vulnerable minima/maxima due adoption fixed switch probability bounded magnitude (from -1 1). this paper, we propose hybrid Q-learning sine-cosine- based strategy, called (QLSCA). Within QLSCA, eliminate switching probability. Instead, rely on (based penalty reward mechanism) dynamically identify best operation during runtime. Additionally, integrate two operations (Le´vy flight motion crossover) into QLSCA jumping out enhance solution diversity. To assess its performance, adopt for combinatorial test suite minimization problem. Experimental results reveal that statistically superior with regard size reduction compared recent state-of-the-art strategies, including original SCA, particle swarm generator (PSTG), optimization (APSO) cuckoo strategy (CS) at 95% confidence level. However, concerning comparison discrete (DPSO), there no significant difference in performance On positive note, outperforms DPSO certain configurations 90%