An Optimization-Based Data Search Clustering Approach for Multidimensional Datasets

Multidimensional data refers to datasets featuring with multiple columns, often referred to as features or attributes. The challenge in multidimensional data analysis is that clusters and outliers are often detected based on the dataset's features, which may not align well with ground truth in real-world scenarios (e.g., gene expression data). Efficiency is a critical consideration as optimized clustering algorithms must handle the growing size of multidimensional datasets. In this research paper, we have proposed a Sinusoidal Chaotic and Information Entropy based Elephant-Herding Optimization for Clustering (SCIE_EOC) to data search in multidimensional datasets. The result shows the proposed method shows around 92-95 per cent accuracy for different datasets which is around 5 per cent better than the earlier methods.