476 search results - page 3 / 96 » Support Vector Machines with Example Dependent Costs |
SADM
2010
13 years 4 months ago
2010
: This paper proposes twin prototype support vector machine (TVM), a constant space and sublinear time support vector machine (SVM) algorithm for online learning. TVM achieves its ...
ICDM
2009
IEEE
14 years 13 days ago
2009
IEEE
—A fast online algorithm OnlineSVMR for training Ramp-Loss Support Vector Machines (SVMR s) is proposed. It finds the optimal SVMR for t+1 training examples using SVMR built on t...
NIPS
2004
13 years 7 months ago
2004
The standard approach to the classification of objects is to consider the examples as independent and identically distributed (iid). In many real world settings, however, this ass...
JMLR
2010
13 years 17 days ago
2010
This paper develops algorithms to train support vector machines when training data are distributed across different nodes, and their communication to a centralized processing unit...
ICML
2008
IEEE
14 years 6 months ago
2008
IEEE
We propose a new stopping condition for a Support Vector Machine (SVM) solver which precisely reflects the objective of the Leave-OneOut error computation. The stopping condition ...