129 search results - page 17 / 26 » icml 2008 |
ICML
2008
IEEE
15 years 11 months ago
2008
IEEE
Previous work has shown that the difficulties in learning deep generative or discriminative models can be overcome by an initial unsupervised learning step that maps inputs to use...
ICML
2008
IEEE
15 years 11 months ago
2008
IEEE
For two-class classification, it is common to classify by setting a threshold on class probability estimates, where the threshold is determined by ROC curve analysis. An analog fo...
ICML
2008
IEEE
15 years 11 months ago
2008
IEEE
We introduce confidence-weighted linear classifiers, which add parameter confidence information to linear classifiers. Online learners in this setting update both classifier param...
ICML
2010
IEEE
14 years 11 months ago
2010
IEEE
Optimization problems with a nuclear norm regularization, such as e.g. low norm matrix factorizations, have seen many applications recently. We propose a new approximation algorit...
ICML
2008
IEEE
15 years 11 months ago
2008
IEEE
We consider feature extraction (dimensionality reduction) for compositional data, where the data vectors are constrained to be positive and constant-sum. In real-world problems, t...