Multiple Instance Learning with Generalized Support Vector Machines

Stuart Andrews, Thomas Hofmann, and Ioannis Tsochantaridis, Brown University

In pattern classification it is usually assumed that a training set of patterns along with their class labels is available. Multiple-Instance Learning (MIL) generalizes this problem setting by making weaker assumptions about the labeling information. We propose to generalize Support Vector Machines to take into account such weak labeling of the type found in MIL. Our method is able to identify superior discriminant functions, as is demonstrated in experiments on synthetic and image datasets.

This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.