Ringed seal search for global optimization via a sensitive search model / Younes Saadi
This thesis proposes a nature-inspired metaheuristic algorithm for global optimization. The proposed algorithm, which is called Ringed Seal Search (RSS), is inspired from the movement of the animal ringed seal. The proposed algorithm is characterized by a search model namely the sensitive search mod...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2018
|
| Subjects: | |
| Online Access: | http://studentsrepo.um.edu.my/8666/1/Younes.pdf http://studentsrepo.um.edu.my/8666/6/younes.pdf http://studentsrepo.um.edu.my/8666/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This thesis proposes a nature-inspired metaheuristic algorithm for global optimization. The proposed algorithm, which is called Ringed Seal Search (RSS), is inspired from the movement of the animal ringed seal. The proposed algorithm is characterized by a search model namely the sensitive search model, where the exploitation-exploration is adaptively balanced. The quality of the algorithm is comprehensively evaluated on various standard benchmark test functions using variety of quality metrics and using three baseline algorithms for comparison. The time consumption analysis shows that RSS consumes less time compared to its homologs. This result is compatible with the convergence analysis. The solution quality analysis demonstrates that the convergence speed of RSS obtained better solution quality, which can be interpreted as a mature search. The diversity evaluation shows that the proposed algorithm achieved an optimal diversity values in most of the benchmark test functions. The experimental results show that the proposed algorithm in this thesis improves the global optimization quality in uni-objective and multi-objective environments while the exploitation and exploration are adaptively balanced. Finally, the proposed algorithm is applied on a data clustering case study using seven benchmark datasets to validate and check its ability to solve real optimization problems. The obtained results show that the proposed algorithm can be used for data clustering. |
|---|