A New One Class Classifier Based on Ensemble of Binary Classifiers

Abstract

Modeling the observation domain of the vectors in a dataset is crucial in most practical applications. This is more important in the case of multivariate regression problems since the vectors which are not drawn from the same distribution as the training data can turn an interpolation problem into an extrapolation problem where the uncertainty of the results increases dramatically. The aim of one-class classification methods is to model the observation domain of target vectors when there is no novel data or there are very few novel data. In this paper, we propose a new one-class classification method that can be trained with or without novel data and it can model the observation domain using any binary classification method. Experiments on visual, non-visual and synthetic data show that the proposed method produces more accurate results compared with state-of-art methods. In addition, we show that by adding only $10\%$ of novel data into our training data, the accuracy of the proposed method increases considerably.