An improved algorithm for support vector clustering based on maximum entropy principle and kernel matrix.
Chonghui GuoFang LiPublished in: Expert Syst. Appl. (2011)
Keyphrases
- kernel matrix
- maximum entropy principle
- support vector
- kernel function
- kernel methods
- maximum entropy
- support vectors
- generative model
- ls svm
- support vector machine
- input space
- feature space
- hyperplane
- svm classifier
- low rank
- training examples
- feature selection
- training data
- cross validation
- generalization ability
- radial basis function
- loss function
- training samples
- model selection
- machine learning
- multiple kernel learning
- polynomial kernels
- conditional random fields
- data sets
- em algorithm
- higher order
- hyperparameters
- multi class
- prior knowledge
- image processing