A New Extreme Learning Machine Optimized by Firefly Algorithm

Abstract

Extreme learning machine (ELM) is a new type of feedforward neural network. Compared with traditional single hidden layer feedforward neural networks, ELM executes with higher training speed and produces smaller error. Due to random input weights and hidden biases, ELM might need numerous hidden neurons to achieve a reasonable accuracy. A new ELM learning algorithm, which was optimized by the Firefly Algorithm (FA), was proposed in this paper. FA was used to select the input weights and biases of hidden layer, and then the output weights could be calculated. To test the validity of proposed method, a simulation experiments about the approximation curves of the SINC function was done. The results showed that the proposed algorithm achieved better performance with less hidden neurons than other similar methods.