Structured parallel implementation of Tree Echo State Network model selection.
Marco DaneluttoClaudio GallicchioAlessio MicheliMassimo TorquatiDaniele VirgilioPublished in: PARCO (2015)
Keyphrases
- model selection
- parallel implementation
- echo state networks
- cross validation
- recurrent neural networks
- hyperparameters
- parameter estimation
- bayesian learning
- sample size
- regression model
- parallel computers
- statistical learning
- selection criterion
- motion segmentation
- machine learning
- error estimation
- statistical inference
- model selection criteria
- meta learning
- marginal likelihood
- feature selection
- real world
- generalization error
- variable selection
- distributed memory
- information criterion
- reinforcement learning
- parallel implementations
- parameter determination
- mixture model
- bayesian model selection
- neural network
- graphical models
- bayesian information criterion
- leave one out cross validation
- high dimensional
- learning algorithm
- automatic model selection