A Pinball loss based Extreme Learning Machine for pattern classification

In this paper, we have used the pinball loss function to measure the empirical risk in the ELM model. The use of the pinball loss in the place of hinge loss in the ELM model makes it insensitive to the boundary noises and use the full information of the training set. Further, it has been empirically shown that the proposed Pinball loss based Extreme Learning Machine owns better generalization ability than other ELM models.