The effect of finite sample size on the holdout error probability estimator of homoscedastic multi-class Gaussian classification problems.
Moataz El AyadiKonstantinos N. PlataniotisPublished in: IJCNN (2010)
Keyphrases
- multi class
- maximum likelihood
- support vector machine
- multi class classification
- pairwise
- feature selection
- linear classifiers
- binary classifiers
- least squares
- cross validation
- statistical learning theory
- binary classification
- base classifiers
- maximum a posteriori
- upper bound
- multi task
- sample size
- nearest neighbor
- binary classification problems
- feature vectors
- uniform convergence