ProDAG: Projection-induced variational inference for directed acyclic graphs.
Ryan ThompsonEdwin V. BonillaRobert KohnPublished in: CoRR (2024)
Keyphrases
- directed acyclic graph
- variational inference
- bayesian inference
- probabilistic graphical models
- mixture model
- topic models
- gaussian process
- probabilistic model
- variational methods
- latent dirichlet allocation
- posterior distribution
- random variables
- closed form
- conditional independence
- graphical models
- directed graph
- approximate inference
- factor graphs
- causal models
- exponential family
- exact inference
- hyperparameters
- prior information
- optic flow
- knowledge representation
- model selection
- active learning
- belief propagation
- expectation maximization