Nearly Dimension-Independent Convergence of Mean-Field Black-Box Variational Inference

• URL: http://arxiv.org/abs/2505.21721v1
• Date: Tue, 27 May 2025 20:08:28 GMT
• Title: Nearly Dimension-Independent Convergence of Mean-Field Black-Box Variational Inference
• Authors: Kyurae Kim, Yi-An Ma, Trevor Campbell, Jacob R. Gardner,
• Abstract summary: Black-box variational inference converges at almost dimension-independent rate.<n>We prove that our bound on the gradient variance cannot be improved using only spectral bounds on the Hessian of the target log-density.
• Score: 22.339038971795873
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We prove that, given a mean-field location-scale variational family, black-box variational inference (BBVI) with the reparametrization gradient converges at an almost dimension-independent rate. Specifically, for strongly log-concave and log-smooth targets, the number of iterations for BBVI with a sub-Gaussian family to achieve an objective $\epsilon$-close to the global optimum is $\mathrm{O}(\log d)$, which improves over the $\mathrm{O}(d)$ dependence of full-rank location-scale families. For heavy-tailed families, we provide a weaker $\mathrm{O}(d^{2/k})$ dimension dependence, where $k$ is the number of finite moments. Additionally, if the Hessian of the target log-density is constant, the complexity is free of any explicit dimension dependence. We also prove that our bound on the gradient variance, which is key to our result, cannot be improved using only spectral bounds on the Hessian of the target log-density.

Related papers

• Convergence of Unadjusted Langevin in High Dimensions: Delocalization of Bias [13.642712817536072]
  We show that as the dimension $d$ of the problem increases, the number of iterations required to ensure convergence within a desired error increases. A key technical challenge we address is the lack of a one-step contraction property in the $W_2,ellinfty$ metric to measure convergence.
  arXiv  Detail & Related papers  (2024-08-20T01:24:54Z)
• Learning with Norm Constrained, Over-parameterized, Two-layer Neural Networks [54.177130905659155]
  Recent studies show that a reproducing kernel Hilbert space (RKHS) is not a suitable space to model functions by neural networks. In this paper, we study a suitable function space for over- parameterized two-layer neural networks with bounded norms.
  arXiv  Detail & Related papers  (2024-04-29T15:04:07Z)
• Breaking the Heavy-Tailed Noise Barrier in Stochastic Optimization Problems [56.86067111855056]
  We consider clipped optimization problems with heavy-tailed noise with structured density. We show that it is possible to get faster rates of convergence than $mathcalO(K-(alpha - 1)/alpha)$, when the gradients have finite moments of order. We prove that the resulting estimates have negligible bias and controllable variance.
  arXiv  Detail & Related papers  (2023-11-07T17:39:17Z)
• Curvature-Independent Last-Iterate Convergence for Games on Riemannian Manifolds [77.4346324549323]
  We show that a step size agnostic to the curvature of the manifold achieves a curvature-independent and linear last-iterate convergence rate. To the best of our knowledge, the possibility of curvature-independent rates and/or last-iterate convergence has not been considered before.
  arXiv  Detail & Related papers  (2023-06-29T01:20:44Z)
• Convergence and concentration properties of constant step-size SGD through Markov chains [0.0]
  We consider the optimization of a smooth and strongly convex objective using constant step-size gradient descent (SGD) We show that, for unbiased gradient estimates with mildly controlled variance, the iteration converges to an invariant distribution in total variation distance. All our results are non-asymptotic and their consequences are discussed through a few applications.
  arXiv  Detail & Related papers  (2023-06-20T12:36:28Z)
• Convergence of Adam Under Relaxed Assumptions [72.24779199744954]
  We show that Adam converges to $epsilon$-stationary points with $O(epsilon-4)$ gradient complexity under far more realistic conditions. We also propose a variance-reduced version of Adam with an accelerated gradient complexity of $O(epsilon-3)$.
  arXiv  Detail & Related papers  (2023-04-27T06:27:37Z)
• Quantum state tomography via non-convex Riemannian gradient descent [13.100184125419691]
  In this work, we derive a quantum state tomography scheme that improves the dependence on $kappa$ to the scale. The improvement comes from the application the non-varian gradient (RGD) algorithms. Our theoretical results of extremely fast convergence and nearly optimal error bounds are corroborated by numerical results.
  arXiv  Detail & Related papers  (2022-10-10T14:19:23Z)
• Convergence of Gaussian-smoothed optimal transport distance with sub-gamma distributions and dependent samples [12.77426855794452]
  This paper provides convergence guarantees for estimating the GOT distance under more general settings. A key step in our analysis is to show that the GOT distance is dominated by a family of kernel maximum mean discrepancy distances.
  arXiv  Detail & Related papers  (2021-02-28T04:30:23Z)
• Convergence Rates of Stochastic Gradient Descent under Infinite Noise Variance [14.06947898164194]
  Heavy tails emerge in gradient descent (SGD) in various scenarios. We provide convergence guarantees for SGD under a state-dependent and heavy-tailed noise with a potentially infinite variance. Our results indicate that even under heavy-tailed noise with infinite variance, SGD can converge to the global optimum.
  arXiv  Detail & Related papers  (2021-02-20T13:45:11Z)
• Faster Convergence of Stochastic Gradient Langevin Dynamics for Non-Log-Concave Sampling [110.88857917726276]
  We provide a new convergence analysis of gradient Langevin dynamics (SGLD) for sampling from a class of distributions that can be non-log-concave. At the core of our approach is a novel conductance analysis of SGLD using an auxiliary time-reversible Markov Chain.
  arXiv  Detail & Related papers  (2020-10-19T15:23:18Z)
• A Simple Convergence Proof of Adam and Adagrad [74.24716715922759]
  We show a proof of convergence between the Adam Adagrad and $O(d(N)/st)$ algorithms. Adam converges with the same convergence $O(d(N)/st)$ when used with the default parameters.
  arXiv  Detail & Related papers  (2020-03-05T01:56:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.