On the Influence of Ill-conditioned Regression Matrix on Hyper-parameter Estimators for Kernel-based Regularization Methods.
Yue JuTianshi ChenBi-Qiang MuLennart LjungPublished in: CDC (2020)
Keyphrases
- regularization methods
- hyperparameters
- regularization parameter
- support vector
- model selection
- cross validation
- image restoration
- edge preserving
- regularization method
- inverse problems
- structure tensor
- total variation
- bayesian framework
- kernel methods
- closed form
- support vector machine
- bayesian inference
- sample size
- random sampling
- prior information
- maximum a posteriori
- em algorithm
- feature selection
- incomplete data
- noise level
- kernel function
- convex optimization
- maximum likelihood
- parameter settings
- incremental learning