Hamiltonian Monte Carlo using an adjoint-differentiated Laplace approximation: Bayesian inference for latent Gaussian models and beyond.
Charles C. MargossianAki VehtariDaniel SimpsonRaj AgrawalPublished in: NeurIPS (2020)
Keyphrases
- monte carlo
- bayesian inference
- expectation propagation
- gaussian model
- particle filter
- markovian decision
- sequential monte carlo methods
- probabilistic model
- latent variables
- gaussian mixture model
- prior information
- conditional independence
- markov chain
- hyperparameters
- expectation maximization
- approximate inference
- gaussian process
- object tracking
- image intensity
- gaussian distribution
- density estimation
- mixture model
- particle filtering
- markov chain monte carlo
- skin color
- monte carlo tree search
- bayesian networks
- visual tracking
- closed form
- face detection
- image sequences
- kalman filter
- least squares
- optimal solution
- reinforcement learning