Corpus ID: 88522014

Optimizing The Integrator Step Size for Hamiltonian Monte Carlo

@article{Betancourt2014OptimizingTI,
  title={Optimizing The Integrator Step Size for Hamiltonian Monte Carlo},
  author={M. Betancourt and Simon Byrne and M. Girolami},
  journal={arXiv: Methodology},
  year={2014}
}
Hamiltonian Monte Carlo can provide powerful inference in complex statistical problems, but ultimately its performance is sensitive to various tuning parameters. In this paper we use the underlying geometry of Hamiltonian Monte Carlo to construct a universal optimization criteria for tuning the step size of the symplectic integrator crucial to any implementation of the algorithm as well as diagnostics to monitor for any signs of invalidity. An immediate outcome of this result is that the… Expand
55 Citations

Figures from this paper

Identifying the Optimal Integration Time in Hamiltonian Monte Carlo
  • 21
  • PDF
A Condition Number for Hamiltonian Monte Carlo.
  • 2
  • Highly Influenced
  • PDF
Faster Hamiltonian Monte Carlo by Learning Leapfrog Scale
  • 6
  • PDF
Adaptive Tuning Of Hamiltonian Monte Carlo Within Sequential Monte Carlo
  • 11
  • Highly Influenced
  • PDF
Discontinuous Hamiltonian Monte Carlo for discrete parameters and discontinuous likelihoods
  • 22
  • PDF
Diagnosing Suboptimal Cotangent Disintegrations in Hamiltonian Monte Carlo
  • 21
  • PDF
Exponential Integration for Hamiltonian Monte Carlo
  • 19
  • Highly Influenced
  • PDF
A Conceptual Introduction to Hamiltonian Monte Carlo
  • 441
  • PDF
The Fundamental Incompatibility of Scalable Hamiltonian Monte Carlo and Naive Data Subsampling
  • 42
  • Highly Influenced
  • PDF
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 22 REFERENCES
Optimal tuning of the hybrid Monte Carlo algorithm
  • 228
  • PDF
MCMC Using Hamiltonian Dynamics
  • 1,838
  • PDF
Generalizing the No-U-Turn Sampler to Riemannian Manifolds
  • 32
  • PDF
On the estimation and correction of discretization error in molecular dynamics averages
  • 5
  • PDF
The Geometric Foundations of Hamiltonian Monte Carlo
  • 114
  • PDF
An improved acceptance procedure for the hybrid Monte Carlo algorithm
  • 136
  • PDF
The No-U-turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo
  • 1,974
  • Highly Influential
  • PDF
Hamiltonian Monte Carlo for Hierarchical Models
  • 162
  • PDF
Hybrid Monte Carlo
  • 692
...
1
2
3
...