论文信息 - Drift Severity Metric

Drift Severity Metric

Concept drift in data is usually considered only as abrupt or gradual thus referring to the speed of change. Such simple distinguishing by speed is sufficient for most of the problems, but there might be situations for which a finer representation would be of use. This paper studies further the phenomenon of concept drift and introduces a simple measure which is relevant to the speed and amount of change between different concepts.

[1] E. S. Page. CONTINUOUS INSPECTION SCHEMES , 1954 .

[2] Shai Ben-David,et al. Learning Changing Concepts by Exploiting the Structure of Change , 1996, COLT '96.

[3] GamaJoão,et al. Decision trees for mining data streams , 2006 .

[4] Kenneth O. Stanley. Learning Concept Drift with a Committee of Decision Trees , 2003 .

[5] Albert Bifet,et al. Massive Online Analysis , 2009 .

[6] Xin Yao,et al. Using diversity to handle concept drift in on-line learning , 2009, 2009 International Joint Conference on Neural Networks.

[7] João Gama,et al. Decision trees for mining data streams , 2006, Intell. Data Anal..

[8] João Gama,et al. Learning with Drift Detection , 2004, SBIA.

[9] William Nick Street,et al. A streaming ensemble algorithm (SEA) for large-scale classification , 2001, KDD '01.