Performance of Arithmetic Optimization Algorithm for ELM Tuning Applied to IoT security

Abstract

In this paper, recently proposed arithmetic optimization algorithm for adopted for extreme learning machine tuning with the goal of improving security internet of things systems. One of the greatest challenges with extreme learning machine is finding initial values of weights and biases and determining satisfying number of neurons in the hidden layer for every particular task. To tackle this challenge, which is NP-hard by nature, the arithmetic optimization algorithm was employed. The proposed method was evaluated against ToN IoT Windows 10 dataset for internet of things security for binary classification, and compared to other extreme learning structures evolved by other state-of-the-art metaheuristics under the same experimental conditions. According to experimental findings, arithmetic optimization algorithm is a promising method for tuning extreme learning machine for this particular challenge.