Cross-Validation EM Training for Robust Parameter Estimation.
Takahiro ShinozakiMari OstendorfPublished in: ICASSP (4) (2007)
Keyphrases
- cross validation
- training set
- robust parameter estimation
- model selection
- hyperparameters
- training and test data
- cross validated
- support vector
- generalization error
- variable selection
- grid search
- expectation maximization
- test set
- regularization parameter
- decision trees
- em algorithm
- maximum likelihood
- data sets
- information criterion
- image segmentation
- training examples
- error estimates
- training samples
- mixture model
- probabilistic model
- leave one out cross validation
- training data