Gradient Hyper-parameter Optimization for Manifold Regularization.
Cassiano O. BeckerPaulo A. V. FerreiraPublished in: ICMLA (2) (2013)
Keyphrases
- manifold regularization
- hyperparameters
- semi supervised learning
- semi supervised classification
- model selection
- semi supervised
- parameter settings
- cross validation
- closed form
- sample size
- bayesian inference
- support vector
- unlabeled data
- dimension reduction
- generative model
- random sampling
- em algorithm
- data dependent
- maximum likelihood
- lower bound