Laplacian Hamiltonian Monte Carlo

Abstract

We proposed a Hamiltonian Monte Carlo (HMC) method with Laplace kinetic energy, and demonstrate the connection between slice sampling and proposed HMC method in one-dimensional cases. Based on this connection, one can perform slice sampling using a numerical integrator in an HMC fashion. We provide theoretical analysis on the performance of such sampler in several univariate cases. Furthermore, the proposed approach extends the standard HMC by enabling sampling from discrete distributions. We compared our method with standard HMC on both synthetic and real data, and discuss its limitations and potential improvements.

Cite

Text

Zhang et al. "Laplacian Hamiltonian Monte Carlo." European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2016. doi:10.1007/978-3-319-46128-1_7

Markdown

[Zhang et al. "Laplacian Hamiltonian Monte Carlo." European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2016.](https://mlanthology.org/ecmlpkdd/2016/zhang2016ecmlpkdd-laplacian/) doi:10.1007/978-3-319-46128-1_7

BibTeX

@inproceedings{zhang2016ecmlpkdd-laplacian,
  title     = {{Laplacian Hamiltonian Monte Carlo}},
  author    = {Zhang, Yizhe and Chen, Changyou and Henao, Ricardo and Carin, Lawrence},
  booktitle = {European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases},
  year      = {2016},
  pages     = {98-114},
  doi       = {10.1007/978-3-319-46128-1_7},
  url       = {https://mlanthology.org/ecmlpkdd/2016/zhang2016ecmlpkdd-laplacian/}
}