Skip to Main content Skip to Navigation
Preprints, Working Papers, ...

Improved polytope volume calculations based on Hamiltonian Monte Carlo with boundary reflections and sweet arithmetics

Abstract : Computing the volume of a high dimensional polytope is a fundamental problem in geometry, also connected to the calculation of densities of states in statistical physics, and a central building block of such algorithms is the method used to sample a target probability distribution. This paper studies Hamiltonian Monte Carlo (HMC) with reflections on the boundary of a domain, providing an enhanced alternative to Hit-and-run (HAR) to sample a target distribution restricted to the polytope. We make three contributions. First, we provide a convergence bound, paving the way to more precise mixing time analysis. Second, we present a robust implementation based on multi-precision arithmetic-a mandatory ingredient to guarantee exact predicates and robust constructions. We however allow controlled failures to happen, introducing the Sweeten Exact Geometric Computing (SEGC) paradigm. Third, we use our HMC random walk to perform H-polytope volume calculations, using it as an alternative to HAR within the volume algorithm by Cousins and Vempala. The tests, conducted up to dimension 50, show that the HMC random walk outperforms HAR.
Document type :
Preprints, Working Papers, ...
Complete list of metadatas

https://hal.inria.fr/hal-03048725
Contributor : Sylvain Pion <>
Submitted on : Wednesday, December 9, 2020 - 3:00:57 PM
Last modification on : Tuesday, December 15, 2020 - 3:56:16 AM

File

socg.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-03048725, version 1

Citation

Augustin Chevallier, Sylvain Pion, Frédéric Cazals. Improved polytope volume calculations based on Hamiltonian Monte Carlo with boundary reflections and sweet arithmetics. 2020. ⟨hal-03048725v1⟩

Share

Metrics

Record views

27

Files downloads

25