Posterior-Scaled MPE: Novel Discriminative Training Criteria.
Markus Nußbaum-ThomZoltán TüskeGeorg HeigoldRalf SchlüterHermann NeyPublished in: INTERSPEECH (2012)
Keyphrases
- discriminative training
- posterior probability
- bayesian networks
- hidden markov models
- maximum likelihood
- probabilistic interpretation
- probability distribution
- bayesian framework
- discriminatively trained
- generative model
- belief networks
- minimum classification error
- probabilistic model
- probability density function
- support vector machine
- automatic speech recognition
- random variables
- unsupervised learning
- expectation maximization
- object detection