Comparative study of fuzzy symbiotic organism search variants for pairwise test suite generation
Metaheuristic algorithms have been utilized for the past 30 years as a core in solving complex optimization problems because of their ability to explore (i.e., roaming the entire search space) and exploit (i.e., searching around the neighbourhood). Most of these algorithms rely on parameter control...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English English |
| Published: |
Institution of Engineering and Technology
2022
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/42000/1/Comparative%20study%20of%20fuzzy%20symbiotic%20organism%20search%20variants.pdf http://umpir.ump.edu.my/id/eprint/42000/2/Comparative%20study%20of%20fuzzy%20symbiotic%20organism%20search%20variants%20for%20pairwise%20test%20suite%20generation_ABS.pdf http://umpir.ump.edu.my/id/eprint/42000/ https://doi.org/10.1049/icp.2022.2576 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Metaheuristic algorithms have been utilized for the past 30 years as a core in solving complex optimization problems because of their ability to explore (i.e., roaming the entire search space) and exploit (i.e., searching around the neighbourhood). Most of these algorithms rely on parameter control to balance this exploration and exploitation to find the best solution. However, tuning these parameters is problematic as they are problem-dependent, and an improper tuning of these parameters undesirably increases computational efforts and yields sub-optimal solutions. Fuzzy Symbiotic Organism Search (FSOS) is among the latest parameter-less meta-heuristics algorithm created to solve optimization problems by having an adaptive exploration and exploitation based on the search need. As this new algorithm is dependent on a Fuzzy Inference System (FIS), the interest in investigating the fuzzy design choice in FSOS has emerged to make sure the choices of the Fuzzy Inference System in FSOS are capable of solving the general optimization problem without overfitting or underfitting. In this paper, we present the effects of different versions of fuzzy rules in the FSOS Fuzzy Inference System on the performance of FSOS. Experimental results demonstrate that the original FSOS with fuzzy rules that cover most of the antecedent combinations supersedes the other combination by 0.7% (FSOS1) and 0.3% (FSOS2). |
|---|