An efficient high-dimensional gene selection approach based on the Binary Horse Herd Optimization Algorithm for biologicaldata classification

The Horse Herd Optimization Algorithm (HOA) is a new meta-heuristic algorithm inspired by the behaviors of horses of different ages. It was recently introduced to solve complex and high-dimensional problems. This paper proposes a binary version of the HOA, termed the Binary Horse Herd Optimization Algorithm (BHOA), designed to solve discrete problems and select prominent feature subsets. Additionally, this study introduces a novel hybrid feature selection framework that combines the BHOA with a Minimum Redundancy Maximum Relevance (MRMR) filter method. This hybrid approach, more computationally efficient, yields a beneficial subset of relevant and informative features. Recognizing feature selection as a binary problem, we have applied a new Transfer Function (TF), named the X-shape TF, which converts continuous problems into binary search spaces. Moreover, the Support Vector Machine (SVM) is employed to evaluate the efficiency of the proposed method on ten microarray datasets: Lymphoma, Prostate, Brain-1, DLBCL, SRBCT, Leukemia, Ovarian, Colon, Lung, and MLL. Compared to other state-of-the-art methods, such as the Gray Wolf (GW), Particle Swarm Optimization (PSO), and Genetic Algorithm (GA), our proposed hybrid method (MRMR-BHOA) demonstrates superior performance in terms of accuracy and minimal selected features. Experimental results also show that the X-shaped BHOA approach outperforms other methods.