查看文章

Introduction: The aim of the feature selection is to find the most important information from a set of given features. It has been used widely in recent years. The set of features primarily has associated features that influence classification performance and increase training time. So, the feature selection is necessary to remove inappropriate features and improve classification performance. In this paper, a nature-inspired feature selection method is proposed based on the behavior of grasshoppers. Grasshopper optimization algorithm is expressed for continuous problems as standard.

Methods: In this paper, continuous grasshopper optimization algorithm is converted into binary by using a nonlinear mapping function. Also, the binary grasshopper optimization algorithm will be converted into a powerful algorithm by combining and retaining exploration and exploitation.

Result: the dataset taken from the UCI repository is used. By comparing the proposed method with GA, Binary Bat Algorithm (BBA), we indicate that it has had better performance than previous methods in terms of accuracy, reduction rate, precision, and computational time.

Conclusion: In this paper, the feature selection problem is presented by the binary grasshopper optimization algorithm. The results are compared at different algorithms. It was concluded that the binary grasshopper optimization algorithm is the best method for feature selection