Context-dependent meta-control for reinforcement learning using a Dirichlet process Gaussian mixture model.
Dongjae KimSang Wan LeePublished in: BCI (2018)
Keyphrases
- gaussian mixture model
- context dependent
- mixture model
- dirichlet process
- reinforcement learning
- em algorithm
- expectation maximization
- feature vectors
- generative model
- probabilistic model
- gaussian distribution
- language model
- low level
- multi task learning
- variational inference
- maximum likelihood
- feature space
- model selection
- natural language
- probability density function
- learning algorithm
- high level
- unsupervised learning
- information retrieval
- support vector
- bayesian framework
- parameter estimation
- density estimation
- markov random field
- text mining
- multi class