Efficient Ensemble Classification for Multi-Label Data Streams with Concept Drift
Most existing multi-label data streams classification methods focus on extending single-label streams classification approaches to multi-label cases, without considering the special characteristics of multi-label stream data, such as label dependency, concept drift, and recurrent concepts. Motivated...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2019-04-01
|
Series: | Information |
Subjects: | |
Online Access: | https://www.mdpi.com/2078-2489/10/5/158 |
id |
doaj-40e44c4cf91a430b9b55fa6137e86e7d |
---|---|
record_format |
Article |
spelling |
doaj-40e44c4cf91a430b9b55fa6137e86e7d2020-11-25T01:36:36ZengMDPI AGInformation2078-24892019-04-0110515810.3390/info10050158info10050158Efficient Ensemble Classification for Multi-Label Data Streams with Concept DriftYange Sun0Han Shao1Shasha Wang2School of Computer and Information Technology, Xinyang Normal University, Xinyang 464000, ChinaSchool of Computer and Information Technology, Xinyang Normal University, Xinyang 464000, ChinaSchool of Computer and Information Technology, Xinyang Normal University, Xinyang 464000, ChinaMost existing multi-label data streams classification methods focus on extending single-label streams classification approaches to multi-label cases, without considering the special characteristics of multi-label stream data, such as label dependency, concept drift, and recurrent concepts. Motivated by these challenges, we devise an efficient ensemble paradigm for multi-label data streams classification. The algorithm deploys a novel change detection based on Jensen−Shannon divergence to identify different kinds of concept drift in data streams. Moreover, our method tries to consider label dependency by pruning away infrequent label combinations to enhance classification performance. Empirical results on both synthetic and real-world datasets have demonstrated its effectiveness.https://www.mdpi.com/2078-2489/10/5/158data streamsmulti-labelconcept driftensemble classificationlabel dependency |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Yange Sun Han Shao Shasha Wang |
spellingShingle |
Yange Sun Han Shao Shasha Wang Efficient Ensemble Classification for Multi-Label Data Streams with Concept Drift Information data streams multi-label concept drift ensemble classification label dependency |
author_facet |
Yange Sun Han Shao Shasha Wang |
author_sort |
Yange Sun |
title |
Efficient Ensemble Classification for Multi-Label Data Streams with Concept Drift |
title_short |
Efficient Ensemble Classification for Multi-Label Data Streams with Concept Drift |
title_full |
Efficient Ensemble Classification for Multi-Label Data Streams with Concept Drift |
title_fullStr |
Efficient Ensemble Classification for Multi-Label Data Streams with Concept Drift |
title_full_unstemmed |
Efficient Ensemble Classification for Multi-Label Data Streams with Concept Drift |
title_sort |
efficient ensemble classification for multi-label data streams with concept drift |
publisher |
MDPI AG |
series |
Information |
issn |
2078-2489 |
publishDate |
2019-04-01 |
description |
Most existing multi-label data streams classification methods focus on extending single-label streams classification approaches to multi-label cases, without considering the special characteristics of multi-label stream data, such as label dependency, concept drift, and recurrent concepts. Motivated by these challenges, we devise an efficient ensemble paradigm for multi-label data streams classification. The algorithm deploys a novel change detection based on Jensen−Shannon divergence to identify different kinds of concept drift in data streams. Moreover, our method tries to consider label dependency by pruning away infrequent label combinations to enhance classification performance. Empirical results on both synthetic and real-world datasets have demonstrated its effectiveness. |
topic |
data streams multi-label concept drift ensemble classification label dependency |
url |
https://www.mdpi.com/2078-2489/10/5/158 |
work_keys_str_mv |
AT yangesun efficientensembleclassificationformultilabeldatastreamswithconceptdrift AT hanshao efficientensembleclassificationformultilabeldatastreamswithconceptdrift AT shashawang efficientensembleclassificationformultilabeldatastreamswithconceptdrift |
_version_ |
1725062032729833472 |