Key instance detection in multi-instance learning
The goal of traditional multi-instance learning (MIL) is to predict the labels of the bags, whereas in many real applications, it is desirable to get the instance labels, especially the labels of key instances that trigger the bag labels, in addition to getting bag labels. Such a problem has been...
Saved in:
Main Authors: | , , |
---|---|
Other Authors: | |
Format: | Conference or Workshop Item |
Language: | English |
Published: |
2014
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/106222 http://hdl.handle.net/10220/23974 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-106222 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-1062222020-05-28T07:18:03Z Key instance detection in multi-instance learning Liu, Guoqing Wu, Jianxin Zhou, Zhi-Hua School of Computer Engineering Asian Conference on Machine Learning (4th : 2012 : Singapore) DRNTU::Engineering::Computer science and engineering The goal of traditional multi-instance learning (MIL) is to predict the labels of the bags, whereas in many real applications, it is desirable to get the instance labels, especially the labels of key instances that trigger the bag labels, in addition to getting bag labels. Such a problem has been largely unexplored before. In this paper, we formulate the Key Instance Detection (KID) problem, and propose a voting framework (VF) solution to KID. The key of VF is to exploit the relationship among instances, represented by a citer kNN graph. This graph is different from commonly used nearest neighbor graphs, but is suitable for KID. Experiments validate the effectiveness of VF for KID. Additionally, VF also outperforms state-of-the-art MIL approaches on the performance of bag label prediction. Published version 2014-10-09T07:36:54Z 2019-12-06T22:06:48Z 2014-10-09T07:36:54Z 2019-12-06T22:06:48Z 2012 2012 Conference Paper Liu, G., Wu, J., & Zhou, Z.-H. (2012). Key instance detection in multi-instance learning. JMLR: Workshop and Conference Proceedings: Asian Conference on Machine Learning, 253-268. https://hdl.handle.net/10356/106222 http://hdl.handle.net/10220/23974 en © 2012 The Author(s). This paper was published in JMLR: Workshop and Conference Proceedings: Asian Conference on Machine Learning and is made available as an electronic reprint (preprint) with permission of the Author(s). One print or electronic copy may be made for personal use only. Systematic or multiple reproduction, distribution to multiple locations via electronic or other means, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper is prohibited and is subject to penalties under law. 16 p. application/pdf |
institution |
Nanyang Technological University |
building |
NTU Library |
country |
Singapore |
collection |
DR-NTU |
language |
English |
topic |
DRNTU::Engineering::Computer science and engineering |
spellingShingle |
DRNTU::Engineering::Computer science and engineering Liu, Guoqing Wu, Jianxin Zhou, Zhi-Hua Key instance detection in multi-instance learning |
description |
The goal of traditional multi-instance learning (MIL) is to predict the labels of the bags,
whereas in many real applications, it is desirable to get the instance labels, especially the
labels of key instances that trigger the bag labels, in addition to getting bag labels. Such a
problem has been largely unexplored before. In this paper, we formulate the Key Instance
Detection (KID) problem, and propose a voting framework (VF) solution to KID. The key
of VF is to exploit the relationship among instances, represented by a citer kNN graph. This
graph is different from commonly used nearest neighbor graphs, but is suitable for KID.
Experiments validate the effectiveness of VF for KID. Additionally, VF also outperforms state-of-the-art MIL approaches on the performance of bag label prediction. |
author2 |
School of Computer Engineering |
author_facet |
School of Computer Engineering Liu, Guoqing Wu, Jianxin Zhou, Zhi-Hua |
format |
Conference or Workshop Item |
author |
Liu, Guoqing Wu, Jianxin Zhou, Zhi-Hua |
author_sort |
Liu, Guoqing |
title |
Key instance detection in multi-instance learning |
title_short |
Key instance detection in multi-instance learning |
title_full |
Key instance detection in multi-instance learning |
title_fullStr |
Key instance detection in multi-instance learning |
title_full_unstemmed |
Key instance detection in multi-instance learning |
title_sort |
key instance detection in multi-instance learning |
publishDate |
2014 |
url |
https://hdl.handle.net/10356/106222 http://hdl.handle.net/10220/23974 |
_version_ |
1681057168814505984 |