多策略人工蜂鸟算法的特征选择方法OA

A multi-strategy artificial hummingbird algorithm(MAHA)is proposed to address the issues of low accuracy and susceptibility to local optima in the iterative process of the artificial hummingbird algorithm.Kent chaotic mapping is introduced during the initialization phase of hummingbird population to improve population diversity and distribution uniformity,thereby en-hancing the algorithm's exploration capability.A differential mutation strategy is applied to update the positions of hummingbirds so that they can escape from local optima.Levy flight improvement algorithm is adopted to enhance the quality of candidate solutions,thus improving convergence speed and accuracy.The MAHA and classical intelligent optimization algorithms were tested on 9 benchmark functions and 6 UCI datasets,and the results shows that MAHA exhibited excellent performance in both accuracy and convergence speed.The feature selection method based on MAHA significantly outperforms other comparative algorithms in classi-fication accuracy and performance.