Performance Analysis of Single and Multi-Objective Metaheuristic Algorithms for Data Clustering Problems

Data clustering is a fundamental approach in unsupervised machine learning that can be utilized to discover patterns with unlabelled data. The variety of algorithms have been adopted to determine the more accurate solutions for data clustering problems. In literature various single and multi-objective metaheuristics algorithms are explored to handle the data clustering problems as well as capable to deal with the high-dimensional and complex data. In this paper, an attempt is made to evaluate and compare the single-objective and multi-objective metaheuristics algorithms for data clustering. Apart from this, this paper also explores the algorithmic design, performance metrics, and empirical outcomes. The simulation results demonstrated that single-objective approaches may be sufficient for simpler clustering tasks, but the multi-objective algorithms often provide superior results in terms of cluster validity, flexibility, and application to real-world, multi-faceted clustering problems.