95 search results - page 18 / 19 » A Comparative Study of Two Novel Predictor Set Scoring Metho... |
BMCBI
2010
13 years 5 months ago
2010
Background: This paper introduces the notion of optimizing different norms in the dual problem of support vector machines with multiple kernels. The selection of norms yields diff...
WEBI
2010
Springer
13 years 3 months ago
2010
Springer
High-dimensional data usually incur learning deficiencies and computational difficulties. We present a novel semi-supervised dimensionality reduction technique that embeds high-dim...
RECOMB
2003
Springer
14 years 6 months ago
2003
Springer
Recent research has demonstrated quite convincingly that accurate cancer diagnosis can be achieved by constructing classifiers that are designed to compare the gene expression pro...
CIKM
2007
Springer
13 years 12 months ago
2007
Springer
Heterogeneous entities or objects are very common and are usually interrelated with each other in many scenarios. For example, typical Web search activities involve multiple types...
ECCV
2006
Springer
14 years 7 months ago
2006
Springer
The paper introduces a new framework for feature learning in classification motivated by information theory. We first systematically study the information structure and present a n...