6 problems
- 0 votes0 replies0 views
The non-extendability conjecture for generalized Monte Carlo in higher dimensions
Let generalized Monte Carlo (GMC) be the discrete-time reflective Hamiltonian Markov chain described in the surrounding construction. An extension to continuous-time dynamics would…
- 0 votes0 replies0 views
Conjecture that the main total-variation convergence theorem extends to HMC
HMC convergence conjecture. Theorem should hold for the HMC algorithm as well. The conjecture proposes that the paper's main non-asymptotic convergence result in total variation ex…
- 0 votes0 replies1 view
Conjecture on the optimality of the HMC time-step restriction for mixing rates
Let denote the HMC integration time, subject to the time-step restriction discussed in the analysis. Time-step optimality conjecture. For many problems of interest, this restri…
- 0 votes0 replies0 views
Generic-gradient-stiffness conjecture for MCMC methods
Consider gradient-based Markov chain Monte Carlo methods, including Hamiltonian Monte Carlo (HMC), applied to rough energy landscapes. Generic-gradient-stiffness conjecture. The we…
- 0 votes0 replies1 view
HMC has no asymptotic improvement over random walk for ball-like multimodal targets
Let be a mixture distribution of the form … where are probability densities, are fixed weights satisfying…
- 0 votes0 replies1 view
Geometric ergodicity conjecture for randomized Hamiltonian Monte Carlo
Let Algorithm denote the approximate randomized Hamiltonian Monte Carlo algorithm, and let the potential energy function and the reversible, volume-preserving integrator satisfy su…