Equivalence of Generative and Log-Linear Models.
Georg HeigoldHermann NeyPatrick LehnenTobias GassRalf SchlüterPublished in: IEEE Trans. Speech Audio Process. (2011)
Keyphrases
- log linear models
- discriminative training
- log linear
- mixture model
- generative model
- hidden markov models
- efficient learning
- conditional random fields
- unsupervised learning
- probabilistic interpretation
- information retrieval
- probability distribution
- markov random field
- semi supervised learning
- latent variables
- probabilistic modeling