We formulate a principle for classification with the knowledge of the marginal distribution over the data points (unlabeled data). The principle is cast in terms of Tikhonov styl...
We propose a novel method of dimensionality reduction for supervised learning. Given a regression or classification problem in which we wish to predict a variable Y from an expla...
Kenji Fukumizu, Francis R. Bach, Michael I. Jordan
— In many applications of supervised learning, the conditional average of the target variables is not sufficient for prediction. The dependencies between the explanatory variabl...