Double Randomized Underdamped Langevin with Dimension-Independent Convergence Guarantee

Authors: Yuanshi Liu, Cong Fang, Tong Zhang

NeurIPS 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical This paper focuses on the high-dimensional sampling of log-concave distributions with composite structures: p (dx) exp( g(x) f(x))dx. We develop a double randomization technique, which leads to a fast underdamped Langevin algorithm with a dimension-independent convergence guarantee. We prove that the algorithm enjoys an overall e O (tr(H))1/3 ϵ2/3 iteration complexity to reach an ϵ-tolerated sample whose distribution p admits W2(p, p ) ϵ. Here, H is an upper bound of the Hessian matrices for f and does not explicitly depend on dimension d. For the posterior sampling over linear models with normalized data, we show a clear superiority of convergence rate which is dimension-free and outperforms the previous best-known results by a d1/3 factor. The analysis to achieve a faster convergence rate brings new insights into high-dimensional sampling.
Researcher Affiliation Academia School of Intelligence Science and Technology, Peking University Department of Computer Science & Engineering, the Hong Kong University of Science and Technology {liu_yuanshi, fangcong}@pku.edu.cn, tongzhang@tongzhang-ml.org
Pseudocode Yes Algorithm 1 Double-Randomized Underdamped Langevin (DRUL)
Open Source Code No The paper does not provide any statement or link regarding the availability of open-source code for the described methodology.
Open Datasets Yes As shown in Figure 1(a), the Gram matrix of the MNIST dataset, which is also the Hessian of Bayesian ridge linear regression, has rapidly decreasing eigenvalues.
Dataset Splits No The paper is theoretical and discusses convergence rates and complexities. It does not describe an experimental setup with training, validation, and test dataset splits.
Hardware Specification No The paper is theoretical and focuses on algorithm design and convergence analysis. It does not describe specific experimental setups or hardware used to obtain empirical results.
Software Dependencies No The paper focuses on theoretical contributions and does not provide details on specific software dependencies with version numbers for reproducibility.
Experiment Setup No The paper is theoretical and focuses on algorithm design and convergence analysis. It does not provide specific details about an experimental setup, such as hyperparameters or training configurations.