Search

Feature selection is a required preprocess stage in most of the data mining tasks. This paper presents an improved Harris hawks optimization (HHO) to find high-quality solutions for global optimization and feature selection tasks. This method is an efficient optimizer inspired by the behaviors of Harris' hawks, which try to catch the rabbits. In some cases, the original version tends to stagnate to the local optimum solutions. Hence, a novel HHO called IHHO is proposed by embedding the salp swarm algorithm (SSA) into the original HHO to improve the search ability of the optimizer and expand the application fields. The update stage in the HHO optimizer, which is performed to update each hawk, is divided into three phases: adjusting population based on SSA to generate SSA-based population, generating hybrid individuals according to SSA-based individual and HHO-based individual, and updating search agent in the light of greedy selection and HHO's mechanisms. A large group of experiments on many functions is carried out to investigate the efficacy of the proposed optimizer. Based on the overall results, the proposed IHHO can provide a faster convergence speed and maintain a better balance between exploration and exploitation. Moreover, according to the proposed continuous IHHO, a more stable binary IHHO is also constructed as a wrapper-based feature selection (FS) approach. We compare the resulting binary IHHO with other FS methods using well-known benchmark datasets provided by UCI. The experimental results reveal that the proposed IHHO has better accuracy rates over other compared wrapper FS methods. Overall research and analysis confirm the improvement in IHHO because of the suitable exploration capability of SSA. [ABSTRACT FROM AUTHOR]

As an optimization paradigm, Salp Swarm Algorithm (SSA) outperforms various population-based optimizers in the perspective of the accuracy of obtained solutions and convergence rate. However, SSA gets stuck into sub-optimal solutions and degrades accuracy while solving the complex optimization problems. To relieve these shortcomings, a modified version of the SSA is proposed in the present work, which tries to establish a more stable equilibrium between the exploration and exploitation cores. This method utilizes two different strategies called opposition-based learning and levy-flight (LVF) search. The algorithm is named m-SSA, and its validation is performed on a well-known set of 23 classical benchmark problems. To observe the strength of the proposed method on the scalability of the test problems, the dimension of these problems is varied from 50 to 1000. Furthermore, the proposed m-SSA is also used to solve some real engineering optimization problems. The analysis of results through various statistical measures, convergence rate, and statistical analysis ensures the effectiveness of the proposed strategies integrated with the m-SSA. The comparison of the m-SSA with the conventional SSA, variants of SSA and some other state-of-the-art algorithms illustrate its enhanced search efficiency. [ABSTRACT FROM AUTHOR]