Sparse Online Co-regularization Using Conjugate Functions

In this paper, we propose a novel sparse online co-regularization framework for multiview semi-supervised learning, which concerns using a small portion of the arrived training examples to represent predictors during the online learning process. This framework makes use of Fenchel conjugates to perform sparse online co-regularization process in the dual function. The use of tolerance function enforces sparsity. Detailed experiments on artificial and real world data sets verify the utility of our approaches. This paper paves a way to the design and analysis of sparse online co-regularization algorithms.