An Improved Gray Wolf Optimization (IGWO) and its Linkage to K-Mean for using in Data Clustering
Ali Al-lami, Adel Ghazikhani, Hussein Al-kaabi
The k-mean algorithm remains one of the most well-known and widespread clustering algorithms,whose initial centers are chosen randomly and while being optimally placed, their application can be easily implemented. The meta-heuristic algorithmcan provide data clustering with the optimal solution. This algorithm can also minimize the issue of local minimums. The present study aimsat improvingthe k-mean algorithm’s accuracy with use of combined and meta-heuristic techniques.Hence, this paperaddresses an algorithm (Improved Gray Wolf OptimizationK-mean). The optimized form of improved gray wolf is employed for automatically detecting the clusters’ number and obtaining the optimal solution as K-mean clustering outcomes and initial K-mean clustering centers. The results revealed that the proposed method bears a lesspercentage of errorcompared to the existing methods and reduces by 12%. Additionally, the aggregate distance of intra-cluster was also decreased.
Volume: Volume 24
Issues: Issue 3
Keywords: Data Clustering, meta-heuristic algorithm, Gray Wolf Optimization (GWO), k-mean.