Random Centroid Selection for K-means Clustering: A Proposed Algorithm for Improving Clustering Results
Arghyadeep Sen, Manjusha Pandey, Krishna Chakravarty · 2020 International Conference on Computer Science, Engineering and Applications (ICCSEA) · 2020
Primary motivation towards this study was to obtain better clustering results from K-means clustering algorithm. Several studies had provided relevant findings showing normal clustering algorithm and scope of improvements regarding clustering accuracy. With the aim to increase clustering accuracy, a conceptual notion of Genetic Algorithm is utilized in K-means clustering algorithm. Depending on the Genetic algorithm concepts, an improved clustering technique is proposed in this study for obtaining more accurate and more precise clustering outcomes. The contribution from this study could be essential in terms of topic-modelled data and clustering text documents.