Dynamic Clustering Scheme for Evolving Data Streams Based on Improved STRAP

Loading...
Thumbnail Image
Access rights
openAccess
Journal Title
Journal ISSN
Volume Title
A1 Alkuperäisartikkeli tieteellisessä aikakauslehdessä
This publication is imported from Aalto University research portal.
View publication in the Research portal
View/Open full text file from the Research portal
Date
2018-09-07
Major/Subject
Mcode
Degree programme
Language
en
Pages
10
46157-46166
Series
IEEE Access, Volume 6
Abstract
A key problem within data mining is clustering of data streams. Most existing algorithms for data stream clustering are based on quite restrictive models for the cluster dynamics. In an attempt to overcome the limitations of existing methods, we propose a novel data stream clustering method, which we refer to as improved streaming affinity propagation (ISTRAP). The ISTRAP is based on an integrated evolution detection framework which ensures that new emerging clusters are recognized timely. Moreover, within ISTRAP, outdated clusters are removed and recurrent clusters are efficiently detected rather than being treated as novel clusters. The proposed ISTRAP is non-parametric in the sense of not requiring any prior information about the number or the centers of clusters. The effectiveness of ISTRAP is evaluated using numerical experiments.
Description
Keywords
data stream clustering, evolving data streams, affinity propagation (AP), on-line clustering
Other note
Citation
Sui, J, Liu, Z, Jung, A, Liu, L & Li, X 2018, ' Dynamic Clustering Scheme for Evolving Data Streams Based on Improved STRAP ', IEEE Access, vol. 6, pp. 46157-46166 . https://doi.org/10.1109/ACCESS.2018.2864553