Skip to main content

Research Repository

Advanced Search

Burst detection-based selective classifier resetting.

Wares, Scott; Isaacs, John; Elyan, Eyad

Authors

Scott Wares



Abstract

Concept drift detection algorithms have historically been faithful to the aged architecture of forcefully resetting the base classifiers for each detected drift. This approach prevents underlying classifiers becoming outdated as the distribution of a data stream shifts from one concept to another. In situations where both concept drift and temporal dependence are present within a data stream, forced resetting can cause complications in classifier evaluation. Resetting the base classifier too frequently when temporal dependence is present can cause classifier performance to appear successful, when in fact this is misleading. In this research, a novel architectural method for determining base classifier resets, Burst Detection-based Selective Classifier Resetting (BD-SCR), is presented. BD-SCR statistically monitors changes in the temporal dependence of a data stream to determine if a base classifier should be reset for detected drifts. The experimental process compares the predictive performance of state-of-the-art drift detectors in comparison to the "No-Change" detector using BD-SCR to inform and control the resetting decision. Results show that BD-SCR effectively reduces the negative impact of temporal dependence during concept drift detection through a clear negation in the performance of the "No-Change" detector, but is capable of maintaining the predictive performance of state-of-the-art drift detection methods.

Citation

WARES, S., ISAACS, J. and ELYAN, E. 2021. Burst detection-based selective classifier resetting. Journal of information and knowledge management [online], 20(2), article 2150027. Available from: https://doi.org/10.1142/S0219649221500271

Journal Article Type Article
Acceptance Date Jan 22, 2021
Online Publication Date Apr 23, 2021
Publication Date Jun 30, 2021
Deposit Date May 6, 2021
Publicly Available Date Apr 24, 2022
Journal Journal of information and knowledge management
Print ISSN 0219-6492
Electronic ISSN 1793-6926
Publisher World Scientific Publishing
Peer Reviewed Peer Reviewed
Volume 20
Issue 2
Article Number 2150027
DOI https://doi.org/10.1142/s0219649221500271
Keywords Data streaming; Concept drift; Temporal dependence
Public URL https://rgu-repository.worktribe.com/output/1328791