Hyper-parameter Selection for Sparse LS-SVM via Minimization of its Localized Generalization Error.
Binbin SunWing W. Y. NgDaniel S. YeungPatrick P. K. ChanPublished in: Int. J. Wavelets Multiresolution Inf. Process. (2013)
Keyphrases
- hyperparameters
- generalization error
- ls svm
- cross validation
- model selection
- sample size
- binary classification
- variable selection
- active learning
- parameter estimation
- random sampling
- training set
- support vector
- high dimensional
- machine learning
- neural network
- upper bound
- closed form
- feature selection
- maximum likelihood
- incremental learning
- least squares
- error rate
- noise level
- feature extraction
- decision trees