Evolutionary Optimisation of Kernel and Hyper-Parameters for SVM.
Laura DiosanAlexandrina RogozanJean-Pierre PécuchetPublished in: MCO (2008)
Keyphrases
- hyperparameters
- support vector
- kernel function
- kernel parameters
- gaussian processes
- genetic algorithm
- gaussian process
- cross validation
- decision function
- support vector machine
- svm classifier
- ls svm
- linear classifiers
- feature selection
- model selection
- grid search
- bayesian methods
- generalization ability
- prior information
- bayesian inference
- sample size
- closed form
- bayesian framework
- random sampling
- probabilistic model
- conjugate priors
- noise level
- maximum a posteriori
- error rate
- active learning
- feature space
- decision trees
- image processing