arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2212.08549stat.COastro-ph.IMhep-lathep-th
收藏

微正则哈密顿蒙特卡洛

Microcanonical Hamiltonian Monte Carlo

University of California at Berkeley(加州大学伯克利分校);Stanford University(斯坦福大学);Lawrence Berkeley National Laboratory(劳伦斯伯克利国家实验室)

Jakob Robnik, G. Bruno De Luca, Eva Silverstein, Uroš Seljak

AI总结:

本文提出微正则哈密顿蒙特卡洛(MCHMC),通过固定能量哈密顿动力学和能量守恒的动量反弹实现遍历性,并开发了连续方向保持反弹的欠阻尼朗之万变体(MCLMC),在多个基准问题上性能优于NUTS HMC一个数量级以上。

摘要:

我们发展了微正则哈密顿蒙特卡洛(MCHMC),这是一类遵循固定能量哈密顿动力学的模型,与遵循不同能量水平正则分布的哈密顿蒙特卡洛(HMC)形成对比。MCHMC调整哈密顿函数,使得动量变量上常能量曲面的均匀分布的边缘分布给出期望的目标分布。我们证明MCHMC需要偶尔的能量守恒台球式动量反弹以实现遍历性,类似于HMC中的动量重采样。我们将反弹概念推广到连续版本,在每一步进行部分方向保持反弹,这给出了具有非高斯噪声的能量守恒欠阻尼朗之万动力学(MCLMC)。MCHMC和MCLMC在条件数和维度上表现出有利的缩放性质。我们开发了一种高效的超参数调整方案,该方案在几个标准基准问题上实现了高性能并始终优于NUTS HMC,在某些情况下性能提升超过一个数量级。

英文摘要:

We develop Microcanonical Hamiltonian Monte Carlo (MCHMC), a class of models which follow a fixed energy Hamiltonian dynamics, in contrast to Hamiltonian Monte Carlo (HMC), which follows canonical distribution with different energy levels. MCHMC tunes the Hamiltonian function such that the marginal of the uniform distribution on the constant-energy-surface over the momentum variables gives the desired target distribution. We show that MCHMC requires occasional energy conserving billiard-like momentum bounces for ergodicity, analogous to momentum resampling in HMC. We generalize the concept of bounces to a continuous version with partial direction preserving bounces at every step, which gives an energy conserving underdamped Langevin-like dynamics with non-Gaussian noise (MCLMC). MCHMC and MCLMC exhibit favorable scalings with condition number and dimensionality. We develop an efficient hyperparameter tuning scheme that achieves high performance and consistently outperforms NUTS HMC on several standard benchmark problems, in some cases by more than an order of magnitude.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑