Identifying uncertainty regions in Support Vector Machines using geometric margin and convex hulls.
Calin VoichitaPurvesh KhatriSorin DraghiciPublished in: IJCNN (2008)
Keyphrases
- convex hull
- support vector
- large margin classifiers
- hyperplane
- learning machines
- maximum margin
- computational geometry
- geometric structure
- soft margin
- multi class
- generalization ability
- kernel function
- closest points
- point sets
- support vector machine
- kernel methods
- binary classification
- loss function
- generalization bounds
- extreme points
- convex concave
- svm classifier
- input image
- standard svm
- face recognition
- image sequences