.png){kind=icon}
返回
Stochastic automatic differentiation for Monte Carlo processes
.png){kind=icon}
PRE
AI
DOI：10.1016/j.cpc.2024.109396.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Monte Carlo methods represent a cornerstone of computer science. They allow sampling high dimensional distribution functions in an efficient way. In this paper we consider the extension of Automatic Differentiation (AD) techniques to Monte Carlo processes, addressing the problem of obtaining derivatives (and in general, the Taylor series) of expectation values. Borrowing ideas from the lattice field theory community, we examine two approaches. One is based on reweighting while the other represents an extension of the Hamiltonian approach typically used by the Hybrid Monte Carlo (HMC) and similar algorithms. We show that the Hamiltonian approach can be understood as a change of variables of the reweighting approach, resulting in much reduced variances of the coefficients of the Taylor series. This work opens the door to finding other variance reduction techniques for derivatives of expectation values.
Keyword:
Automatic differentiation
Monte Carlo
Lattice field theory
Optimization
Bayesian inference
期刊
IF:
3.4
论文数:
1.2W
被引数:
3.7W
机构
引用论文
The relationship between blood glycosylated haemoglobin and home capillary blood glucose levels in diabetics.png)
Diabetologia
OAAI
DiabetologiaIF0
OAAI分享收藏Synchronization of Wind Trubine Generators Against an Infinite Bus Under Gusting Wind Conditions
PREAI
分享收藏
查看更多内容