Bayesian updating and marginal likelihood estimation by cross entropy based importance sampling.
Michael EngelOindrila KanjilalIason PapaioannouDaniel StraubPublished in: J. Comput. Phys. (2023)
Keyphrases
- importance sampling
- marginal likelihood
- approximate inference
- graphical models
- model selection
- posterior distribution
- hyperparameters
- monte carlo
- gaussian process
- parameter estimation
- belief propagation
- expectation propagation
- markov chain monte carlo
- probabilistic inference
- message passing
- bayesian networks
- closed form
- kalman filter
- exponential family
- particle filter
- information criterion
- conditional random fields
- latent variables
- markov chain
- particle filtering
- support vector
- pairwise
- random sampling
- noise level
- incremental learning
- bayesian framework
- maximum a posteriori
- mixture model
- em algorithm
- maximum likelihood
- information extraction
- active learning