An extreme learning machine (ELM) is a single layer feed-forward neural network alike learning system, whose connections from the input layer to the hidden layer are randomly generated, while the connections from the hidden layer to the output layer are learned through linear learning methods.