Nonlinear Stochastic Gradient Descent and Heavy-tailed Noise: A Unified Framework and High-probability Guarantees

• URL: http://arxiv.org/abs/2410.13954v1
• Date: Thu, 17 Oct 2024 18:25:28 GMT
• Title: Nonlinear Stochastic Gradient Descent and Heavy-tailed Noise: A Unified Framework and High-probability Guarantees
• Authors: Aleksandar Armacki, Shuhua Yu, Pranay Sharma, Gauri Joshi, Dragana Bajovic, Dusan Jakovetic, Soummya Kar,
• Abstract summary: We study high-probability convergence in online learning, in the presence of heavy-tailed noise. Compared to state-of-the-art, who only consider clipping and require noise with bounded $p$-th moments, we provide guarantees for a broad class of nonlinearities.
• Score: 56.80920351680438
• License:
• Abstract: We study high-probability convergence in online learning, in the presence of heavy-tailed noise. To combat the heavy tails, a general framework of nonlinear SGD methods is considered, subsuming several popular nonlinearities like sign, quantization, component-wise and joint clipping. In our work the nonlinearity is treated in a black-box manner, allowing us to establish unified guarantees for a broad range of nonlinear methods. For symmetric noise and non-convex costs we establish convergence of gradient norm-squared, at a rate $\widetilde{\mathcal{O}}(t^{-1/4})$, while for the last iterate of strongly convex costs we establish convergence to the population optima, at a rate $\mathcal{O}(t^{-\zeta})$, where $\zeta \in (0,1)$ depends on noise and problem parameters. Further, if the noise is a (biased) mixture of symmetric and non-symmetric components, we show convergence to a neighbourhood of stationarity, whose size depends on the mixture coefficient, nonlinearity and noise. Compared to state-of-the-art, who only consider clipping and require unbiased noise with bounded $p$-th moments, $p \in (1,2]$, we provide guarantees for a broad class of nonlinearities, without any assumptions on noise moments. While the rate exponents in state-of-the-art depend on noise moments and vanish as $p \rightarrow 1$, our exponents are constant and strictly better whenever $p < 6/5$ for non-convex and $p < 8/7$ for strongly convex costs. Experiments validate our theory, demonstrating noise symmetry in real-life settings and showing that clipping is not always the optimal nonlinearity, further underlining the value of a general framework.

Related papers

• Large Deviations and Improved Mean-squared Error Rates of Nonlinear SGD: Heavy-tailed Noise and Power of Symmetry [47.653744900375855]
  We study large deviations and mean-squared error (MSE) guarantees a general framework of nonlinear convex gradient methods in the online setting. We provide strong results for broad bound step-sizes in the presence of heavy noise symmetric density function.
  arXiv  Detail & Related papers  (2024-10-21T04:50:57Z)
• 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)
• High-probability Convergence Bounds for Nonlinear Stochastic Gradient Descent Under Heavy-tailed Noise [59.25598762373543]
  We show that wetailed high-prob convergence guarantees of learning on streaming data in the presence of heavy-tailed noise. We demonstrate analytically and that $ta$ can be used to the preferred choice of setting for a given problem.
  arXiv  Detail & Related papers  (2023-10-28T18:53:41Z)
• Distributed Extra-gradient with Optimal Complexity and Communication Guarantees [60.571030754252824]
  We consider monotone variational inequality (VI) problems in multi-GPU settings where multiple processors/workers/clients have access to local dual vectors. Extra-gradient, which is a de facto algorithm for monotone VI problems, has not been designed to be communication-efficient. We propose a quantized generalized extra-gradient (Q-GenX), which is an unbiased and adaptive compression method tailored to solve VIs.
  arXiv  Detail & Related papers  (2023-08-17T21:15:04Z)
• High Probability Bounds for a Class of Nonconvex Algorithms with AdaGrad Stepsize [55.0090961425708]
  We propose a new, simplified high probability analysis of AdaGrad for smooth, non- probability problems. We present our analysis in a modular way and obtain a complementary $mathcal O (1 / TT)$ convergence rate in the deterministic setting. To the best of our knowledge, this is the first high probability for AdaGrad with a truly adaptive scheme, i.e., completely oblivious to the knowledge of smoothness.
  arXiv  Detail & Related papers  (2022-04-06T13:50:33Z)
• Nonlinear gradient mappings and stochastic optimization: A general framework with applications to heavy-tail noise [11.768495184175052]
  We introduce a general framework for nonlinear gradient descent scenarios when gradient noise exhibits heavy tails. We show that for a nonlinearity with bounded outputs and for the gradient noise that may not have finite moments of order greater than one, the nonlinear SGD converges to zero at rate$O(/tzeta)$, $zeta in (0,1)$. Experiments show that, while our framework is more general than existing studies of SGD under heavy-tail noise, several easy-to-implement nonlinearities from our framework are competitive with state of the art alternatives on real data sets
  arXiv  Detail & Related papers  (2022-04-06T06:05:52Z)
• 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)

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.