Cost-sensitive online classification
Both cost-sensitive classification and online learning have been studied extensively in data mining and machine learning communities, respectively. It is a bit surprising that there was very limited comprehensive study for addressing an important intersecting problem, that is, cost-sensitive online...
Main Authors: | , , |
---|---|
Other Authors: | |
Format: | Conference Paper |
Language: | English |
Published: |
2013
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/99920 http://hdl.handle.net/10220/13025 |
_version_ | 1811696252658647040 |
---|---|
author | Hoi, Steven C. H. Wang, Jialei. Zhao, Peilin. |
author2 | School of Computer Engineering |
author_facet | School of Computer Engineering Hoi, Steven C. H. Wang, Jialei. Zhao, Peilin. |
author_sort | Hoi, Steven C. H. |
collection | NTU |
description | Both cost-sensitive classification and online learning have been studied extensively in data mining and machine learning communities, respectively. It is a bit surprising that there was very limited comprehensive study for addressing an important intersecting problem, that is, cost-sensitive online classification. In this paper, we formally study this problem, and propose a new framework for cost-sensitive online classification by exploiting the idea of online gradient descent techniques. Based on the framework, we propose a family of cost-sensitive online classification algorithms, which are designed to directly optimize two well-known cost-sensitive measures: (i) maximization of weighted sum of sensitivity and specificity, and (ii) minimization of weighted misclassification cost. We analyze the theoretical bounds of the cost-sensitive measures made by the proposed algorithms, and extensively examine their empirical performance on a variety of cost-sensitive online classification tasks. |
first_indexed | 2024-10-01T07:36:25Z |
format | Conference Paper |
id | ntu-10356/99920 |
institution | Nanyang Technological University |
language | English |
last_indexed | 2024-10-01T07:36:25Z |
publishDate | 2013 |
record_format | dspace |
spelling | ntu-10356/999202020-05-28T07:17:26Z Cost-sensitive online classification Hoi, Steven C. H. Wang, Jialei. Zhao, Peilin. School of Computer Engineering IEEE International Conference on Data Mining (12th : 2012 : Brussels, Belgium) DRNTU::Engineering::Computer science and engineering Both cost-sensitive classification and online learning have been studied extensively in data mining and machine learning communities, respectively. It is a bit surprising that there was very limited comprehensive study for addressing an important intersecting problem, that is, cost-sensitive online classification. In this paper, we formally study this problem, and propose a new framework for cost-sensitive online classification by exploiting the idea of online gradient descent techniques. Based on the framework, we propose a family of cost-sensitive online classification algorithms, which are designed to directly optimize two well-known cost-sensitive measures: (i) maximization of weighted sum of sensitivity and specificity, and (ii) minimization of weighted misclassification cost. We analyze the theoretical bounds of the cost-sensitive measures made by the proposed algorithms, and extensively examine their empirical performance on a variety of cost-sensitive online classification tasks. 2013-08-06T02:57:17Z 2019-12-06T20:13:38Z 2013-08-06T02:57:17Z 2019-12-06T20:13:38Z 2012 2012 Conference Paper https://hdl.handle.net/10356/99920 http://hdl.handle.net/10220/13025 10.1109/ICDM.2012.116 en |
spellingShingle | DRNTU::Engineering::Computer science and engineering Hoi, Steven C. H. Wang, Jialei. Zhao, Peilin. Cost-sensitive online classification |
title | Cost-sensitive online classification |
title_full | Cost-sensitive online classification |
title_fullStr | Cost-sensitive online classification |
title_full_unstemmed | Cost-sensitive online classification |
title_short | Cost-sensitive online classification |
title_sort | cost sensitive online classification |
topic | DRNTU::Engineering::Computer science and engineering |
url | https://hdl.handle.net/10356/99920 http://hdl.handle.net/10220/13025 |
work_keys_str_mv | AT hoistevench costsensitiveonlineclassification AT wangjialei costsensitiveonlineclassification AT zhaopeilin costsensitiveonlineclassification |