Partial Optimal Transport with Applications on Positive-Unlabeled Learning

Neural Information Processing Systems 

In this paper, we address the partial Wasserstein and Gromov-Wasserstein problems and propose exact algorithms to solve them.