Asymptotic Log-loss of Prequential Maximum Likelihood Codes
Peter GrünwaldSteven de RooijPublished in: CoRR (2005)
Keyphrases
- maximum likelihood
- rates of convergence
- expectation maximization
- em algorithm
- parameter estimation
- log likelihood
- bayes error
- gaussian mixture model
- error correction
- likelihood function
- cross validated
- worst case
- machine learning
- log likelihood function
- expected values
- asymptotically optimal
- information loss
- gaussian distribution
- hypothesis testing
- exponential family
- large deviations
- error correcting
- error correcting codes
- synthetic aperture radar
- image segmentation
- neural network