Chemical Reaction-Based Optimization Algorithm for Solving Clustering Problems

Abstract

Clustering is an important data analysis and well-recognized technique in the field of data mining. Since past few decades, it has attained a wide awareness from research community having theoretical as well as practical views. It is an unsupervised classification technique, and there is no need for training and testing of data objects. This technique can split the data objects into different clusters using a distance criterion. Literature survey has revealed the use of this technique in diverse research domains [3, 4, 7, 8, 18, 19, 21] . In recent years, a number of heuristics algorithms have been reported to determine the optimal solution for clustering problems. These algorithms have been derived from the heuristics, natural phenomena, swarms behavior, insects, animals, etc. Some physics-based algorithms (such as charged system search algorithm, galaxy-based algorithm, black hole optimization algorithm, magnetic charged system search algorithm) have also been used to find out the optimal solution for optimization problems [6, 9, 10, 16]. Some other algorithms like CSS [11], MCSS [12], BH [6], and CSO [13, 14] have also been applied successfully to solve the clustering problems. With time, the