DeepAI AI Chat
Log In Sign Up

Magnetic Hamiltonian Monte Carlo

07/10/2016
by   Nilesh Tripuraneni, et al.
0

Hamiltonian Monte Carlo (HMC) exploits Hamiltonian dynamics to construct efficient proposals for Markov chain Monte Carlo (MCMC). In this paper, we present a generalization of HMC which exploits non-canonical Hamiltonian dynamics. We refer to this algorithm as magnetic HMC, since in 3 dimensions a subset of the dynamics map onto the mechanics of a charged particle coupled to a magnetic field. We establish a theoretical basis for the use of non-canonical Hamiltonian dynamics in MCMC, and construct a symplectic, leapfrog-like integrator allowing for the implementation of magnetic HMC. Finally, we exhibit several examples where these non-canonical dynamics can lead to improved mixing of magnetic HMC relative to ordinary HMC.

READ FULL TEXT

page 1

page 2

page 3

page 4

10/15/2020

Magnetic Manifold Hamiltonian Monte Carlo

Markov chain Monte Carlo (MCMC) algorithms offer various strategies for ...
08/18/2020

Non-Canonical Hamiltonian Monte Carlo

Hamiltonian Monte Carlo is typically based on the assumption of an under...
05/04/2020

Connecting the Dots: Towards Continuous Time Hamiltonian Monte Carlo

Continuous time Hamiltonian Monte Carlo is introduced, as a powerful alt...
09/29/2020

Couplings for Andersen Dynamics

Andersen dynamics is a standard method for molecular simulations, and a ...
04/14/2021

Zaionc paradox revisited

Canonical expressions are representative of implicative propositions upt...
11/25/2017

Generalizing Hamiltonian Monte Carlo with Neural Networks

We present a general-purpose method to train Markov chain Monte Carlo ke...
11/09/2020

Symmetrically processed splitting integrators for enhanced Hamiltonian Monte Carlo sampling

We construct integrators to be used in Hamiltonian (or Hybrid) Monte Car...