801 search results - page 133 / 161 » The Inefficiency of Batch Training for Large Training Sets |
PKDD
2001
Springer
14 years 1 months ago
2001
Springer
Large amount of available information does not necessarily imply that induction algorithms must use all this information. Samples often provide the same accuracy with less computat...
FGR
2000
IEEE
14 years 1 months ago
2000
IEEE
Image variability due to changes in pose and illumination can seriously impair object recognition. This paper presents appearance-based methods which, unlike previous appearance-b...
INFOCOM
2000
IEEE
14 years 1 months ago
2000
IEEE
— Trainet, a new scheme to extend MPLS (Multi Protocol Label Switching) is presented. The scheme works much like the subway system in a large metropolitan area. Each (unidirectio...
ACIVS
2007
Springer
14 years 25 days ago
2007
Springer
We present a patch-based algorithm for the purpose of object classification in video surveillance. Within detected regions-of-interest (ROIs) of moving objects in the scene, a feat...
CEC
2007
IEEE
14 years 24 days ago
2007
IEEE
Abstract-- Hypernetworks consist of a large number of hyperedges that represent higher-order features sampled from training patterns. Evolutionary algorithms have been used as a me...