|
gptkbp:instanceOf
|
Markov chain
sampling algorithm
|
|
gptkbp:advantage
|
improves mixing
reduces random walk behavior
|
|
gptkbp:alsoKnownAs
|
gptkb:Hybrid_Monte_Carlo
|
|
gptkbp:application
|
computational physics
statistical modeling
Bayesian hierarchical models
|
|
gptkbp:category
|
gptkb:Monte_Carlo_methods
gptkb:Computational_statistics
Bayesian statistics
|
|
gptkbp:field
|
gptkb:machine_learning
statistics
Bayesian inference
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Hamiltonian Monte Carlo
|
|
gptkbp:implementedIn
|
gptkb:PyMC
gptkb:TensorFlow_Probability
gptkb:Stan
gptkb:Turing.jl
|
|
gptkbp:introduced
|
gptkb:Anthony_D._Kennedy
gptkb:Brian_J._Pendleton
gptkb:Duncan_Roweth
gptkb:Simon_Duane
|
|
gptkbp:introducedIn
|
1987
|
|
gptkbp:purpose
|
efficient sampling from complex probability distributions
|
|
gptkbp:relatedTo
|
gptkb:Metropolis-Hastings_algorithm
gptkb:No-U-Turn_Sampler
|
|
gptkbp:requires
|
gradient of log probability
|
|
gptkbp:uses
|
Hamiltonian dynamics
leapfrog integrator
|
|
gptkbp:bfsParent
|
gptkb:Markov_chain
|
|
gptkbp:bfsLayer
|
5
|