Multi-objective clustering analysis using educational system algorithm

Data clustering is an unsupervised learning tool which is used to segment a dataset into homogeneous groups based on similarity and dissimilarity metrics. Traditional clustering algorithms often consider a basic assumption on the clustering structure and optimize it by adopting a suitable objective function corresponding to the use of classical or evolutionary methods. These algorithms act poorly when there are no assumptions about data. Multi-objective clustering, in which objective functions are optimized simultaneously, it will be a high-performance alternative in such a situation. In this research, a clustering algorithm is presented based on the multi-objective optimization educational system algorithm, and then its efficiency is evaluated and is compared with other clustering algorithms. Experiments have shown that this algorithm is more efficient and more accurate than other same algorithms.