2193 search results - page 184 / 439 » Properties of Support Vector Machines |
191
click to vote
IJPRAI
2010
15 years 5 months ago
2010
: AUC-SVM directly maximizes the area under the ROC curve (AUC) through minimizing its hinge loss relaxation, and the decision function is determined by those support vector sample...
165
click to vote
CIVR
2008
Springer
15 years 8 months ago
2008
Springer
In this paper, we propose two ways of improving image classification based on bag-of-words representation [25]. Two shortcomings of this representation are the loss of the spatial...
172
click to vote
SAC
2006
ACM
16 years 13 days ago
2006
ACM
Traditional Data Mining and Knowledge Discovery algorithms assume free access to data, either at a centralized location or in federated form. Increasingly, privacy and security co...
180
click to vote
PR
2007
15 years 6 months ago
2007
This article focusses on the problems of feature extraction and the recognition of handwritten digits. A trainable feature extractor based on the LeNet5 convolutional neural netwo...
166
click to vote
ICML
2007
IEEE
16 years 7 months ago
2007
IEEE
Although support vector machines (SVMs) for binary classification give rise to a decision rule that only relies on a subset of the training data points (support vectors), it will ...