LS-SVM Hyperparameter Selection with a Nonparametric Noise Estimator.
Amaury LendasseYongnan JiNima ReyhaniMichel VerleysenPublished in: ICANN (2) (2005)
Keyphrases
- ls svm
- hyperparameters
- least squares
- least squares support vector machine
- noise level
- maximum a posteriori
- maximum likelihood
- cross validation
- random sampling
- noise reduction
- variable selection
- model selection
- rbf kernel
- noisy images
- parameter estimation
- gaussian process
- bayesian framework
- gaussian processes
- support vector
- input variables
- markov random field
- training data
- support vector machine
- fuzzy logic
- image reconstruction
- generalization ability
- em algorithm
- prediction model
- missing data
- probabilistic model
- closed form