Other formats:
BibTeX
LaTeX
RIS
@article{1110133,
author = {Gama, João and Kosina, Petr},
article_location = {London},
article_number = {3},
doi = {http://dx.doi.org/10.1007/s10115-013-0654-6},
keywords = {Data streams; Concept drift; Meta-learning; Recurrent concepts},
language = {eng},
issn = {0219-1377},
journal = {Knowledge and Information Systems},
title = {Recurrent concepts in data streams classification},
url = {http://dx.doi.org/10.1007/s10115-013-0654-6},
volume = {40},
year = {2013}
}
TY - JOUR
ID - 1110133
AU - Gama, João - Kosina, Petr
PY - 2013
TI - Recurrent concepts in data streams classification
JF - Knowledge and Information Systems
VL - 40
IS - 3
SP - 1-19
EP - 1-19
PB - Springer-Verlag
SN - 02191377
KW - Data streams
KW - Concept drift
KW - Meta-learning
KW - Recurrent concepts
UR - http://dx.doi.org/10.1007/s10115-013-0654-6
N2 - This work addresses the problem of mining data streams generated in dynamic environments where the distribution underlying the observations may change over time. We present a system that monitors the evolution of the learning process. The system is able to self-diagnose degradations of this process, using change detection mechanisms, and self-repair the decision models. The system uses meta-learning techniques that characterize the domain of applicability of previously learned models. The meta-learner can detect recurrence of contexts, using unlabeled examples, and take pro-active actions by activating previously learned models. The experimental evaluation on three text mining problems demonstrates the main advantages of the proposed system: it provides information about the recurrence of concepts and rapidly adapts decision models when drift occurs.
ER -
GAMA, João and Petr KOSINA. Recurrent concepts in data streams classification. \textit{Knowledge and Information Systems}. London: Springer-Verlag, 2013, vol.~40, No~3, p.~1-19. ISSN~0219-1377. Available from: https://dx.doi.org/10.1007/s10115-013-0654-6.
|
