Improved Whale Optimization Algorithm For Clustering
Main Article Content
Abstract
Clustering,inthefieldofdatamining,isdefinedastheprocessofgroupingsimilardatapoints.Nature-inspired algorithmsareusedinclusteringtoavoidprematureconvergenceintolocaloptima.Nature-inspiredalgorithms such as cuckoo search, firefly algorithm, bat algorithm, and flower pollination algorithm are defined as algorithms that emulate animals’ behavior in nature under varied circumstances. One such algorithm is the Whale Optimization Algorithm (WOA), inspired by the humpback whales’ bubble-net hunting strategy. Although WOA is observed to outperform several other nature-inspired algorithms, it suffers from exploration-exploitation imbalance and trapping in local optima. This paper proposes an improved Whale Optimization Algorithm with optimized hyperparameters determined using the Grid Search Algorithm to overcome the aforementioned. The proposed work is seen to outperform the existingWOA.
Article Details
All articles published in NVEO are licensed under Copyright Creative Commons Attribution-NonCommercial 4.0 International License.