278 search results - page 33 / 56 » Strong Separation of Learning Classes |
150
Voted
ICML
2000
IEEE
16 years 3 months ago
2000
IEEE
If all features causing heterogeneity were observed, a mixture of experts approach (Jacobs et al., 1991) is likely to be superior to using a single model. When unobserved or very n...
124
Voted
BMVC
2010
15 years 19 days ago
2010
Off-line trained class-specific object detectors are designed to detect any instance of the class in a given image or video sequence. In the context of object tracking, however, o...
106
Voted
ICPR
2010
IEEE
15 years 17 days ago
2010
IEEE
Unlike most previous manifold-based data classification algorithms assume that all the data points are on a single manifold, we expect that data from different classes may reside ...
101
Voted
CAINE
2008
15 years 4 months ago
2008
This paper presents a novel method for learning from a labeled dataset to accurately classify unknown data. The recursive algorithm, termed Recursive Hyperspheric Classification, ...
108
Voted
ML
2010
ACM
15 years 1 months ago
2010
ACM
When only a small number of labeled samples are available, supervised dimensionality reduction methods tend to perform poorly due to overfitting. In such cases, unlabeled samples ...