.png){kind=icon}
.png){kind=icon}
Return
An ensemble multi-swarm teaching-learning-based optimization algorithm for function optimization and image segmentation
.png){kind=icon}
PRE
AI
DOI：10.1016/j.asoc.2022.109653.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Intelligent optimization algorithms are widely utilized to deal with complex optimization problems in various areas. However, a single intelligent optimization algorithm cannot handle well more and more complex optimization problems. The ensemble strategy can integrate several different operators and algorithms by using some appropriate strategies and maybe obtain better optimization performance. In this paper, a new ensemble multi-swarm method based on teaching-learning-based optimization (EMTLBO) was proposed by integrating three different algorithms including the original teaching-learning-based optimization algorithm, its variant with neighborhood search and the variant with differential evolution. In EMTLBO, a new evaluating mechanism based on the fitness-based and diversity-based metrics (FDEM) for each sub-swarm was proposed to evaluate the optimization performance after a continuous generation interval. Moreover, an algorithm matching mechanism based on ranking for sub-swarms (AMM) is adapted to re-divide the population into three sub -swarms and match a suitable algorithm for each sub-swarm so as to increase the whole optimization performance. Furthermore, the experimental results on CEC2014 and CEC2017 test suits verify the feasibility and optimization performance of EMTLBO. Finally, the proposed algorithm is extended to optimize the segmentation thresholds of images and the segmentation performances on different benchmark images show that EMTLBO has good performance in most cases.(c) 2022 Elsevier B.V. All rights reserved.
Keywords:
Ensemble strategy
Teaching-learning-based optimization
Function optimization
Image segmentation
Journal
IF:
6.6
Papers:
1.4W
Citations:
4.8W