Namib beetle optimization algorithm: A new meta-heuristic method for feature selection and dimension reduction

Today, large amounts of data are generated in various applications such as smart cities and social networks, and their processing requires a lot of time. One of the methods of processing data types and reducing computational time on data is the use of dimension reduction methods. Reducing dimensions is a problem with the optimization approach and meta-heuristic methods can be used to solve it. Namib beetles are an example of intelligent insects and creatures in nature that use an interesting strategy to survive and collect water in the desert. In this article, the behavior of Namib beetles has been used to collect water in the desert to model the Namib beetle optimization (NBO) algorithm. In the second phase of a binary version, this algorithm is used to select features and reduce dimensions. Experiments on CEC functions show that the proposed method has fewer errors than the DE, BBO, SHO, WOA, GOA, and HHO algorithms. In large dimensions such as 200, 500, and 1000 dimensions, the NBO algorithm of meta-heuristic algorithms such as HHO and WOA has a better rank in the optimal calculation of benchmark functions. Experiments show that the proposed algorithm has a greater ability to reduce dimensions and feature selection than similar meta-heuristic algorithms. In 87.5% of the experiments, the proposed method reduces the data space more than other compared methods.