330 search results - page 20 / 66 » Learning Concept Importance Using a Weighted Dependence Mode... |
KDD
1995
ACM
13 years 11 months ago
1995
ACM
Manydata mining algorithms developed recently are based on inductive learning methods. Very few are based on similarity-based learning. However, similarity-based learning accrues ...
COMAD
2009
13 years 8 months ago
2009
Mining evolving data streams for concept drifts has gained importance in applications like customer behavior analysis, network intrusion detection, credit card fraud detection. Se...
JETAI
1998
13 years 7 months ago
1998
Graphical structures such as Bayesian networks or Markov networks are very useful tools for representing irrelevance or independency relationships, and they may be used to e cientl...
ICS
2010
Tsinghua U.
14 years 10 days ago
2010
Tsinghua U.
The emergence of multicore processors has increased the need for simple parallel programming models usable by nonexperts. The ability to specify subparts of a bigger data structur...
ICML
2006
IEEE
14 years 8 months ago
2006
IEEE
When constructing a classifier from labeled data, it is important not to assign too much weight to any single input feature, in order to increase the robustness of the classifier....