Reducing label complexity by learning from bags

Sivan Sabato, Nathan Srebro, Naftali Tishby

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

3 Scopus citations

Abstract

We consider a supervised learning setting in which the main cost of learning is the number of training labels and one can obtain a single label for a bag of examples, indicating only if a positive example exists in the bag, as in Multi- Instance Learning. We thus propose to create a training sample of bags, and to use the obtained labels to learn to classify individual examples. We provide a theoretical analysis showing how to select the bag size as a function of the problem parameters, and prove that if the original labels are distributed unevenly, the number of required labels drops considerably when learning from bags. We demonstrate that finding a low-error separating hyperplane from bags is feasible in this setting using a simple iterative procedure similar to latent SVM. Experiments on synthetic and real data sets demonstrate the success of the approach.

Original languageEnglish
Title of host publicationProceedings of the 13th International Conference on Artificial Intelligence and Statistics (AISTATS)
Pages685-692
Number of pages8
Volume9
StatePublished - 2010
Externally publishedYes
Event13th International Conference on Artificial Intelligence and Statistics, AISTATS 2010 - Sardinia, Italy
Duration: 13 May 201015 May 2010

Publication series

NameJournal of Machine Learning Research
PublisherMicrotome Publishing
ISSN (Print)1532-4435

Conference

Conference13th International Conference on Artificial Intelligence and Statistics, AISTATS 2010
Country/TerritoryItaly
CitySardinia
Period13/05/1015/05/10

Fingerprint

Dive into the research topics of 'Reducing label complexity by learning from bags'. Together they form a unique fingerprint.

Cite this