Premium
Grid‐based high performance ensemble classification for evolving data stream
Author(s) -
Qian Quan,
Xie Mengbo,
Xiao Chaojie,
Zhang Rui
Publication year - 2016
Publication title -
concurrency and computation: practice and experience
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.309
H-Index - 67
eISSN - 1532-0634
pISSN - 1532-0626
DOI - 10.1002/cpe.3898
Subject(s) - computer science , speedup , grid , computation , intrusion detection system , data mining , ensemble learning , raw data , data classification , grid computing , artificial intelligence , machine learning , pattern recognition (psychology) , parallel computing , algorithm , geometry , mathematics , programming language
Summary Ensemble learning is one of the main topics of focus in machine learning research. This paper proposes a novel multi‐thread grid‐based multi‐chunk multi‐level ensemble (GMCE) for data stream classification. In order to improve the learning efficiency, GMCE maps different raw data to multiple grids, represents the feature of the grid by the grid first‐order geometric center, and then classifies data based on the grid. Because this grid mapping method compresses the data size significantly, GMCE can increase both the classification accuracy and the computation efficiency. This method has been tested using public KDDCUP99 intrusion detection competition data and five popular P2P applications data. The results show that the GMCE is better than the original multi‐chunk multi‐level ensemble (MCE) in terms of classification accuracy and operation efficiency. For KDDCUP99 data, GMCE has shown more than 98% classification accuracy and 10 times speedup. And for P2P traffic data, GMCE has improved the classification accuracy by 3–4% and achieved greater than four times speedup. Copyright © 2016 John Wiley & Sons, Ltd.