An Efficient Slime Mould Algorithm Combined With K-Nearest Neighbor for Medical Classification Tasks

Growing science and medical technologies have produced a massive amount of knowledge on different scales biological systems. By processing various amounts data, these will increase the quality disease detection enhance usability health information The integration machine learning in computer-based diagnostic systems facilitates early diseases, enabling more productive treatments prolonged survival rates. slime mould algorithm (SMA) may drawbacks, such as being trapped minimal local regions having an unbalanced exploitation exploration phase. To overcome limitations, this paper proposes ISMA, improved version hybridized with opposition-based (OBL) strategy based k-nearest neighbor (kNN) classifier for classification approach. Opposition-based improves global exploratory ability while avoiding premature convergence. experimental results revealed superiority proposed ISMA–kNN evaluation metrics, including accuracy, sensitivity, specificity, precision, F-score, G-mean, computational time, feature selection (FS) size compared tunicate swarm (TSA), marine predators (MPA), chimp optimization (ChOA), moth–flame (MFO) algorithm, whale (WOA), sine cosine (SCA), original SMA algorithm. Performance tests were run same maximum number function evaluations (FEs) nine UCI benchmark data sets sizes.