34 search results - page 2 / 7 » Online classification of nonstationary data streams |
NN
2006
Springer
13 years 4 months ago
2006
Springer
This paper presents an on-line unsupervised learning mechanism for unlabeled data that are polluted by noise. Using a similarity thresholdbased and a local error-based insertion c...
ICONIP
2009
13 years 2 months ago
2009
In contrast to traditional machine learning algorithms, where all data are available in batch mode, the new paradigm of streaming data poses additional difficulties, since data sam...
JMLR
2010
12 years 11 months ago
2010
Massive Online Analysis (MOA) is a software environment for implementing algorithms and running experiments for online learning from evolving data streams. MOA is designed to deal...
ISI
2008
Springer
13 years 4 months ago
2008
Springer
In this paper, we study the problem of anomaly detection in high-dimensional network streams. We have developed a new technique, called Stream Projected Ouliter deTector (SPOT), t...
SIGMOD
2004
ACM
14 years 4 months ago
2004
ACM
Real-time surveillance systems, network and telecommunication systems, and other dynamic processes often generate tremendous (potentially infinite) volume of stream data. Effectiv...