MIT News has published an article featuring the most recent conference paper by graduate students Charli Fogner and Chiyuan Zhang in the Centor of Brains, Minds and Machines, Hossein Mabahi of the Computer Science & Artificial Intelligence Lab at MIT, Mauricio Araya-Polo of Shell Int'l Exploration & Production, and Tomaso Poggio, the Eugene McDermott Professor in the Brain Sciences and Human Behavior and Director of CBMM. 

Project abstract:

Learning to predict multi-label outputs is challenging, but in many problems there is a natural metric on the outputs that can be used to improve predictions. In this paper we develop a loss function for multi-label learning, based on the Wasserstein distance. The Wasserstein distance provides a natural notion of dissimilarity for probability measures. Although optimizing with respect to the exact Wasserstein distance is costly, recent work has described a regularized approximation that is efficiently computed. We describe efficient learning algorithms based on this regularization, extending the Wasserstein loss from probability measures to unnormalized measures. We also describe a statistical learning bound for the loss and show connections with the total variation norm and the Jaccard index. The Wasserstein loss can encourage smoothness of the predictions with respect to a chosen metric on the output space. We demonstrate this property on a real-data tag prediction problem, using the Yahoo Flickr Creative Commons dataset, achieving superior performance over a baseline that doesn't use the metric.

Please also visit the project's web page at: http://cbcl.mit.edu/wasserstein