A New Hyper-parameters Selection Approach for Support Vector Machines to Predict Time Series.
Yanhua YuJunde SongZhijun RenPublished in: ICPCA/SWS (2012)
Keyphrases
- hyperparameters
- support vector
- cross validation
- learning machines
- model selection
- closed form
- bayesian framework
- gaussian process
- bayesian inference
- posterior distribution
- prior information
- em algorithm
- random sampling
- variational bayes
- sample size
- bayesian methods
- feature selection
- noise level
- incremental learning
- support vector machine
- maximum a posteriori
- maximum likelihood
- ls svm
- data sets
- parameter space
- support vector regression
- incomplete data
- similarity measure
- expectation maximization
- super resolution