Dataset Sources and Download Links for BCSMO

Abstract

This dataset collection contains 43 datasets used to evaluate the proposed Elite-Guided Binary Chaos-Enhanced Starling Murmuration Optimization for High-Dimensional Biomedical Feature Selection approach. The collection consists of: 33 benchmark datasets obtained from the UCI Machine Learning Repository, covering different dataset sizes, dimensionalities, and application domains. 10 high-dimensional microarray gene-expression datasets used specifically for biomedical feature selection and gene selection experiments. The datasets cover a wide range of characteristics, from low-dimensional classification problems to highly dimensional biomedical and gene-expression datasets containing thousands of features/genes. This diversity enables comprehensive evaluation of feature-selection algorithms in terms of search effectiveness, scalability, dimensionality reduction, and classification performance. An accompanying file is provided containing the dataset names, dataset identifiers, number of instances, number of features/genes, and the corresponding source or download links for all 43 datasets. These links allow users to access the original datasets from their respective repositories or sources. Associated Research:Elite-Guided Binary Chaos-Enhanced Starling Murmuration Optimization for High-Dimensional Biomedical Feature Selection