Surrogate Learning - An Approach for Semi-Supervised Classification

2008/09/26 by Veeramachaneni, Sriharsha, Kondadadi, Ravikumar
#FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.0809.4632

Abstract

We consider the task of learning a classifier from the feature space X to the set of classes Y = \0, 1\, when the features can be partitioned into class-conditionally independent feature sets X₁ and X₂. We show the surprising fact that the class-conditional independence can be used to represent the original learning task in terms of 1) learning a classifier from X₂ to X₁ and 2) learning the class-conditional distribution of the feature set X₁. This fact can be exploited for semi-supervised learning because the former task can be accomplished purely from unlabeled samples. We present experimental evaluation of the idea in two real world applications.

Related

• One-Class Semi-Supervised Learning: Detecting Linearly Separable Class by its Mean
• Binary Classification from Multiple Unlabeled Datasets via Surrogate Set Classification
• Projected Estimators for Robust Semi-supervised Classification
• Semi-supervised learning
• Semi-Supervised Learning of Classifiers from a Statistical Perspective: A Brief Review