Extreme Learning Machine Classifier Based on Novel Particle Swarm Optimization Algorithm

Aiming at the problem of low classification accuracy caused by random input weights and hidden layer bias in ELM (Extreme Learning Machines, ELM), a method based on IPSO (Improved Particle Swarm Optimization, IPSO) to optimize the network classification model of Extreme Learning Machines is proposed. Firstly, the nonlinear inertia weight and dynamic acceleration factor are used to balance the exploration and mining of the PSO algorithm; Secondly, the variables of the ELM model are enhanced by IPSO to reduce the influence caused by strong volatility; Finally, simulation results show that compared with SVM, ELM, GA-ELM, IPSO-ELM could effectively improve the precision of classification prediction.

I. Introduction

At present, classification has been widely studied in many fields [1]. Traditional extreme learning machine model has some shortcomings such as strong fluctuation and easy overfitting, and it is often difficult to achieve ideal results for practical complex classification problems. The meta-heuristic intelligent algorithm is dedicated to solving complex problems, PSO algorithm has excellent structure, few variables and is easy to implement, but it tends to fall into local optimum because of its characteristics, which greatly reduces its search performance. Therefore, it is meaningful to explore how to effectively combine the improved PSO algorithm with the extreme learning machine model