6 search results - page 1 / 2 » Mining in Anticipation for Concept Change: Proactive-Reactiv... |
DATAMINE
2006
13 years 5 months ago
2006
KDD
2005
ACM
13 years 11 months ago
2005
ACM
Mining data streams is important in both science and commerce. Two major challenges are (1) the data may grow without limit so that it is difficult to retain a long history; and (...
ICMCS
2006
IEEE
13 years 11 months ago
2006
IEEE
Pattern mining algorithms are often much easier applied than quantitatively assessed. In this paper we address the pattern evaluation problem by looking at both the capability of ...
KDD
2007
ACM
14 years 5 months ago
2007
ACM
In many practical applications, one is interested in generating a ranked list of items using information mined from continuous streams of data. For example, in the context of comp...
ICDM
2008
IEEE
13 years 11 months ago
2008
IEEE
To cope with concept drift, we paired a stable online learner with a reactive one. A stable learner predicts based on all of its experience, whereas a reactive learner predicts ba...