Novel clustering algorithms based on jellyfish optimization with applications in gene expression
Accurate gene expression clustering is crucial for identifying cancer subtypes and elucidating underlying disease mechanisms. This paper proposes an enhanced jellyfish search (EJS) algorithm that extends the standard jellyfish search (JS) metaheuristic through two key innovations: a rearrangement improvement strategy that strengthens global exploration and an integrated local search mechanism that intensifies exploitation around promising solutions. This synergistic design accelerates convergence and consistently yields high-quality clustering partitions. The proposed EJS algorithm is rigorously evaluated on ten real-world benchmark gene expression datasets and compared with several state-of-the-art clustering methods, including variance-based harmony search, genetic algorithm, hybrid particle swarm optimization with enhanced K-means matching, cuckoo search, and differential evolution. Experimental results demonstrate that EJS consistently outperforms all competing approaches. It achieves the best overall Friedman rank (1.00) and reduces the average fitness value by 5.7% compared with the best-performing competitor. Furthermore, EJS obtains superior F-measure, Purity, and Calinski–Harabasz index scores across most datasets, while maintaining linear computational complexity, thereby confirming its ability to generate compact, well-separated, and biologically meaningful clusters. These findings are further validated through statistical significance tests, highlighting the robustness and reliability of the proposed method. These results establish EJS as a robust, efficient, and scalable method for high-quality gene expression clustering.
Authors
- Omar Bendjeghaba (ORCID: https://orcid.org/0000-0003-3717-3039)
- Saida Ishak Boushaki (ORCID: https://orcid.org/0000-0003-3082-2884)
- Noureddine Brakta (ORCID: https://orcid.org/0000-0003-3927-3295)
- Sadjia Lahiani (ORCID: https://orcid.org/0000-0003-0828-6886)
Institutions
- University of Boumerdes (DZ)
Publication Details
- Journal
- International Journal of Electrical Engineering Education
- Published
- 2026-11-18
- DOI
- https://doi.org/10.1177/00207209261508576
- Primary Topic
- Advanced Clustering Algorithms Research
- Type
- article
- Field-Weighted Citation Impact
- 0.00