Large-margin minimum classification error training: A theoretical risk minimization perspective.
Dong YuLi DengXiaodong HeAlex AceroPublished in: Comput. Speech Lang. (2008)
Keyphrases
- minimum classification error
- risk minimization
- discriminative training
- hinge loss
- multi category
- loss function
- hidden markov models
- statistical pattern recognition
- maximum likelihood
- maximum likelihood estimation
- probabilistic interpretation
- classification error
- generalization error
- machine learning
- automatic speech recognition
- co training
- graphical models
- reinforcement learning
- learning algorithm