Firefly algorithm optimized extreme learning machine for hyperspectral image classification

A firefly algorithm (FA) based parameter optimization method for extreme learning machine (ELM) for hyperspectral image classification is proposed. The parameters of regularization coefficient and Gaussian kernel are optimized by firefly algorithm in this method. The experimental results show that the FA-based optimization algorithm can provide the better performance of extreme learning machines for hyperspectral image classification, and it outperforms the popular particle swarm optimization (PSO), genetic algorithm (GA) method.