MILEAGE: Multiple Instance LEArning with Global Embedding

Dan Zhang, Jingrui He, Luo Si, Richard D. Lawrence

Research output: Chapter in Book/Report/Conference proceedingConference contribution

8 Citations (Scopus)

Abstract

Multiple Instance Learning (MIL) generally represents each example as a collection of instances such that the features for local objects can be better captured, whereas traditional methods typically extract a global feature vector for each example as an integral part. However, there is limited research work on investigating which of the two learning scenarios performs better. This paper proposes a novel framework - Multiple Instance LEArning with Global Embedding (MILEAGE), in which the global feature vectors for traditional learning methods are integrated into the MIL setting. Within the proposed framework, a large margin method is formulated to adaptively tune the weights on the two different kinds of feature representations (i.e., global and multiple instance) for each example and trains the classifier simultaneously. An extensive set of experiments are conducted to demonstrate the advantages of the proposed method.

Original languageEnglish (US)
Title of host publication30th International Conference on Machine Learning, ICML 2013
PublisherInternational Machine Learning Society (IMLS)
Pages1119-1127
Number of pages9
EditionPART 2
StatePublished - 2013
Externally publishedYes
Event30th International Conference on Machine Learning, ICML 2013 - Atlanta, GA, United States
Duration: Jun 16 2013Jun 21 2013

Other

Other30th International Conference on Machine Learning, ICML 2013
CountryUnited States
CityAtlanta, GA
Period6/16/136/21/13

Fingerprint

learning
Classifiers
learning method
Experiments
scenario
experiment

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Sociology and Political Science

Cite this

Zhang, D., He, J., Si, L., & Lawrence, R. D. (2013). MILEAGE: Multiple Instance LEArning with Global Embedding. In 30th International Conference on Machine Learning, ICML 2013 (PART 2 ed., pp. 1119-1127). International Machine Learning Society (IMLS).

MILEAGE : Multiple Instance LEArning with Global Embedding. / Zhang, Dan; He, Jingrui; Si, Luo; Lawrence, Richard D.

30th International Conference on Machine Learning, ICML 2013. PART 2. ed. International Machine Learning Society (IMLS), 2013. p. 1119-1127.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Zhang, D, He, J, Si, L & Lawrence, RD 2013, MILEAGE: Multiple Instance LEArning with Global Embedding. in 30th International Conference on Machine Learning, ICML 2013. PART 2 edn, International Machine Learning Society (IMLS), pp. 1119-1127, 30th International Conference on Machine Learning, ICML 2013, Atlanta, GA, United States, 6/16/13.
Zhang D, He J, Si L, Lawrence RD. MILEAGE: Multiple Instance LEArning with Global Embedding. In 30th International Conference on Machine Learning, ICML 2013. PART 2 ed. International Machine Learning Society (IMLS). 2013. p. 1119-1127
Zhang, Dan ; He, Jingrui ; Si, Luo ; Lawrence, Richard D. / MILEAGE : Multiple Instance LEArning with Global Embedding. 30th International Conference on Machine Learning, ICML 2013. PART 2. ed. International Machine Learning Society (IMLS), 2013. pp. 1119-1127
@inproceedings{8f7a810056a5489f86ffbaac6d46b341,
title = "MILEAGE: Multiple Instance LEArning with Global Embedding",
abstract = "Multiple Instance Learning (MIL) generally represents each example as a collection of instances such that the features for local objects can be better captured, whereas traditional methods typically extract a global feature vector for each example as an integral part. However, there is limited research work on investigating which of the two learning scenarios performs better. This paper proposes a novel framework - Multiple Instance LEArning with Global Embedding (MILEAGE), in which the global feature vectors for traditional learning methods are integrated into the MIL setting. Within the proposed framework, a large margin method is formulated to adaptively tune the weights on the two different kinds of feature representations (i.e., global and multiple instance) for each example and trains the classifier simultaneously. An extensive set of experiments are conducted to demonstrate the advantages of the proposed method.",
author = "Dan Zhang and Jingrui He and Luo Si and Lawrence, {Richard D.}",
year = "2013",
language = "English (US)",
pages = "1119--1127",
booktitle = "30th International Conference on Machine Learning, ICML 2013",
publisher = "International Machine Learning Society (IMLS)",
edition = "PART 2",

}

TY - GEN

T1 - MILEAGE

T2 - Multiple Instance LEArning with Global Embedding

AU - Zhang, Dan

AU - He, Jingrui

AU - Si, Luo

AU - Lawrence, Richard D.

PY - 2013

Y1 - 2013

N2 - Multiple Instance Learning (MIL) generally represents each example as a collection of instances such that the features for local objects can be better captured, whereas traditional methods typically extract a global feature vector for each example as an integral part. However, there is limited research work on investigating which of the two learning scenarios performs better. This paper proposes a novel framework - Multiple Instance LEArning with Global Embedding (MILEAGE), in which the global feature vectors for traditional learning methods are integrated into the MIL setting. Within the proposed framework, a large margin method is formulated to adaptively tune the weights on the two different kinds of feature representations (i.e., global and multiple instance) for each example and trains the classifier simultaneously. An extensive set of experiments are conducted to demonstrate the advantages of the proposed method.

AB - Multiple Instance Learning (MIL) generally represents each example as a collection of instances such that the features for local objects can be better captured, whereas traditional methods typically extract a global feature vector for each example as an integral part. However, there is limited research work on investigating which of the two learning scenarios performs better. This paper proposes a novel framework - Multiple Instance LEArning with Global Embedding (MILEAGE), in which the global feature vectors for traditional learning methods are integrated into the MIL setting. Within the proposed framework, a large margin method is formulated to adaptively tune the weights on the two different kinds of feature representations (i.e., global and multiple instance) for each example and trains the classifier simultaneously. An extensive set of experiments are conducted to demonstrate the advantages of the proposed method.

UR - http://www.scopus.com/inward/record.url?scp=84897478893&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84897478893&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84897478893

SP - 1119

EP - 1127

BT - 30th International Conference on Machine Learning, ICML 2013

PB - International Machine Learning Society (IMLS)

ER -