Classification and feature selection for microarray cancer dataset using an improved African vulture optimisation algorithm

The African vulture optimisation algorithm (AVOA) is a recently developed metaheuristic algorithm that imitates the eating and movement patterns of authentic African vultures. AVOA's convergence accuracy and stability, like that of various state-of-the-art metaheuristics, will degrade as optimisation problems get more complicated and variable. Furthermore, the traditional AVOA only searches in one direction, limiting its convergence capacity and causing stagnation at local minima. To address these shortcomings, this work offers LOBL-AVOA, a better version of AVOA hybridised with the lens opposition-based learning (LOBL) technique for classification. The LOBL enhances global exploratory capacity while preventing premature convergence. The suggested LOBL-AVOA findings are compared to traditional AVOA results. The efficacy of LOBL-AVOA is assessed employing six high-dimensional microarray datasets and three distinct classifiers such as support vector machine (SVM), K-nearest-neighbour (KNN) and random forest (RF). As an outcome, the proposed method surpasses traditional AVOA in terms of convergence capability, statistical analysis and classification accuracy.