K-means Clustering Hybridized with Improved Chaos Based Particle Swarm Optimization for High Dimensional Problem

PSO algorithms are gaining considerable attention because of its involvement in swarm evolutionary computing. However, there are still issues in achieving a balance between exploration and exploitation. This work presents a hybridized technique that combines K-means and a revolutionary PSO-based algorithm to tackle these problems. The research introduces Chaos Adaptive Particle Swarm Optimization (CAPSO), which utilizes chaos theory to dynamically modify important parameters and a regulating factor. This approach improves the adaptability of the algorithm and prevents premature convergence of the PSO algorithm. Subsequently, the integration of K-means clustering with CAPSO is employed to expedite convergence and employ chaotic search for enhancing the Optimal Point Globally in High dimensional data. The efficiency of KCAPSO in solving difficult multidimensional problems has been confirmed by experimental findings. KCAPSO exhibits significant advancements compared to traditional K-means, rendering them highly helpful for utilization in diverse optimization endeavors.