Penalized least squares, model selection, convex hull classes and neural nets.
Gerald H. L. CheangAndrew R. BarronPublished in: ESANN (2001)
Keyphrases
- model selection
- convex hull
- neural nets
- least squares
- parameter estimation
- cross validation
- feed forward
- back propagation
- hyperparameters
- variable selection
- sample size
- machine learning
- neural network
- generalization error
- mixture model
- selection criterion
- artificial neural networks
- gaussian process
- feature selection
- unsupervised learning
- marginal likelihood
- learning algorithm
- convex polyhedra
- learning tasks
- bayesian information criterion
- information criterion
- model selection criteria
- optimal solution
- high dimensional
- optical flow
- maximum likelihood