Learning and Inference in Mixed-State Conditionally Heteroskedastic Factor Models Using Viterbi Approximation.
Mohamed SaidaneChristian LavergnePublished in: International Conference on Computational Science (4) (2006)
Keyphrases
- structured prediction
- learning algorithm
- prior knowledge
- learning process
- learning systems
- active learning
- supervised learning
- dynamic bayesian networks
- learning tasks
- probabilistic inference
- learning models
- learning problems
- inference process
- learned models
- inductive inference
- random fields
- knowledge acquisition
- hidden variables
- machine learning
- decision theoretic
- markov models
- probabilistic model
- reinforcement learning
- bayesian networks
- latent variable models