Improved equilibrium optimization based on Levy flight approach for feature selection

In this research, an enhanced variant of equilibrium optimization is proposed to handle feature selection problems. Retrieving the relevant features from high dimensional micro array gene expression data is important. It is mandate to form the way in which the diversified merger of the Levy flight using the feature selection (FS) concept. We have incorporated the Levy flight (LF) approach with the conventional equilibrium optimization (EO) for exclusively FS. In the proposed model, randomization utilizing the Levy flight enhances convergence efficiency greatly by removing local minimum stagnation. The proposed method is tested with six standard micro-array cancer datasets and compared with the conventional algorithms and conventional EO. The results show that the recommended model excels in terms of convergence ability and classification precision in the most of high-dimensional datasets.