The asymptotics of semi-supervised learning in discriminative probabilistic models

Nataliya Sokolovska, Olivier Cappé, François Yvon

Abstract: Semi-supervised learning aims at taking advantage of unlabeled data to improve the efficiency of supervised learning procedures. For discriminative models however, this is a challenging task. In this contribution, we introduce an original methodology for using unlabeled data through the design of a simple semi-supervised objective function. We prove that the corresponding semi-supervised estimator is asymptotically optimal. The practical consequences of this result are discussed for the case of the logistic regression model.