发表机构
Massachusetts Institute of Technology; Yale University; Duke University(麻省理工学院; 耶鲁大学; 杜克大学)
机构由 AI 辅助整理,请以论文原文为准。AI 中文总结
本文通过尖锐维度分析证明,在强凸Langevin采样中,确定性中点方法(如Heun和指数中点)的梯度查询复杂度优于随机中点方法,并给出匹配下界。
AI 中文摘要
我们研究目标分布 $\pi \propto e^{-V}$ 的 Langevin 动力学的确定性和随机中点离散化,其中 $0 \prec \alpha I\preceq\nabla^2V\preceq\beta I$ 且 $\kappa=\beta/\alpha$。为达到 $\sqrt\alpha\\,W_2\leqslant\varepsilon$,我们证明确定性 Heun 方法至多需要 $\widetilde O(\kappa^{4/3}d^{1/3}\varepsilon^{-2/3})$ 次梯度查询,而欠阻尼指数中点方法需要 $\widetilde O(\kappa^{5/4}d^{1/4}\varepsilon^{-1/2})$ 次。证明利用了平稳性下的抵消和 Malliavin 微积分技术进行平滑,优于基于标准耦合的先前上界。在条件数有界时,下界与两种确定性方法的 $d$ 和 $\varepsilon$ 的幂次匹配。对比之下,对于随机中点方法和至少两个网格点的 Poisson 中点方法(包括过阻尼和欠阻尼变体),一个简单的高斯计算得出下界 $d^{1/3}\varepsilon^{-1/3}$,即使从良性初始化开始,要达到 $\varepsilon$-接近的样本也需要该下界。这令人惊讶地表明,在高维中,确定性离散化可以优于其随机对应方法。
英文摘要
We study deterministic and randomized midpoint discretizations of Langevin dynamics for a target $π\propto e^{-V}$, where $0 \prec αI\preceq\nabla^2V\preceqβI$ and $κ=β/α$. To achieve $\sqrtα\,W_2\leqslant\varepsilon$, we show that deterministic Heun uses at most $\widetilde O(κ^{4/3}d^{1/3}\varepsilon^{-2/3})$ gradient queries, and underdamped exponential midpoint uses $\widetilde O(κ^{5/4}d^{1/4}\varepsilon^{-1/2})$. The proofs exploit cancellation at stationarity and smoothing using techniques from Malliavin calculus, outperforming previous upper bounds based on standard couplings. At bounded condition number, a lower bound matches the $d$ and $\varepsilon$ powers of both deterministic methods. To contrast, for the randomized midpoint methods and Poisson midpoint with at least two grid points (both overdamped and underdamped variants), a simple Gaussian calculation yields a lower bound $d^{1/3}\varepsilon^{-1/3}$ to get an $\varepsilon$-close sample despite starting at a benign initialization. This shows surprisingly that in high dimensions, deterministic discretizations can outperform their random counterparts.