Please use this identifier to cite or link to this item:
https://hdl.handle.net/2440/134833
Citations | ||
Scopus | Web of Science® | Altmetric |
---|---|---|
?
|
?
|
Type: | Conference paper |
Title: | Weighted ensemble classification of multi-label data streams |
Author: | Wang, L. Shen, H. Tian, H. |
Citation: | Lecture Notes in Artificial Intelligence, 2017 / Kim, J., Shim, K., Cao, L., Lee, J.G., Lin, X., Moon, Y.S. (ed./s), vol.10235, pp.551-562 |
Publisher: | Springer |
Issue Date: | 2017 |
Series/Report no.: | Lecture Notes in Artificial Intelligence |
ISBN: | 9783319575285 |
ISSN: | 0302-9743 1611-3349 |
Conference Name: | Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD 2017) (23 May 2017 - 26 May 2017 : Jeju, South Korea) |
Editor: | Kim, J. Shim, K. Cao, L. Lee, J.G. Lin, X. Moon, Y.S. |
Statement of Responsibility: | Lulu Wang, Hong Shen, and Hui Tian |
Abstract: | Many real world applications involve classification of multilabel data streams. However, most existing classification models mostly focused on classifying single-label data streams. Learning in multi-label data stream scenarios is more challenging, as the classification systems should be able to consider several properties, such as large data volumes, label correlations and concept drifts. In this paper, we propose an efficient and effective ensemble model for multi-label stream classification based on ML-KNN (Multi-Label KNN) [31] and propose a balance AdjustWeight function to combine the predictions which can efficiently process high-speed multi-label stream data with concept drifts. The empirical results indicate that our approach achieves a high accuracy and low storage cost, and outperforms the existing methods ML-KNN and SMART [14]. |
Keywords: | Multi-label; Data stream; Classification |
Rights: | © Springer International Publishing AG 2017 |
DOI: | 10.1007/978-3-319-57529-2_43 |
Grant ID: | http://purl.org/au-research/grants/arc/DP150104871 |
Published version: | http://www.springer.com/series/1244 |
Appears in Collections: | Computer Science publications |
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.