3556 search results - page 92 / 712 » Can machine learning be secure |
ECML
2006
Springer
14 years 6 days ago
2006
Springer
This paper presents a decomposition method for efficiently constructing 1-norm Support Vector Machines (SVMs). The decomposition algorithm introduced in this paper possesses many d...
ICMLC
2010
Springer
13 years 8 months ago
2010
Springer
: The performance of a statistical machine translation (SMT) system heavily depends on the quantity and quality of the bilingual language resource. However, the pervious work mainl...
ECML
2003
Springer
14 years 3 months ago
2003
Springer
In multi-instance learning, the training set comprises labeled bags that are composed of unlabeled instances, and the task is to predict the labels of unseen bags. Through analyzin...
ISCA
2007
IEEE
14 years 4 months ago
2007
IEEE
Virtualization has become much more important throughout the computer industry both to improve security and to support multiple workloads on the same hardware with effective isola...
NIPS
1998
13 years 11 months ago
1998
The kernel-parameter is one of the few tunable parameters in Support Vector machines, controlling the complexity of the resulting hypothesis. Its choice amounts to model selection...