In nature,biological invasions have attracted attention because of their rapid development and significant ecological impacts.The introduction of populations to search for suitable habitats often has inherent logic,and communication between populations and population expansion also play an important role in this process.A Biological Invasion-Based Feature Selection(BIAFS)algorithm is proposed by exploring the principle of population searches for suitable habitats.In the BIAFS algorithm,the biological invasion process is divided into four stages:population establishment,migration,communication and expansion,and development.During the experimental verification process,the BIAFS algorithm is compared with eight high-performance algorithms using nine datasets.The experimental results show that the Classification Accuracy(CA)and Dimensionality Reduction(DR)rate of the BIAFS algorithm on the seven datasets exceeded those of the comparison algorithm.In addition,a comparative experiment with the fitness standard deviation also confirms the high stability of the BIAFS algorithm,indicating that it can more robustly determine the optimal solution in multiple datasets.These experimental results demonstrate the effectiveness and superiority of the BIAFS algorithm for feature selection tasks.
万方数据
维普
