Revisiting Step-Size Assumptions in Stochastic Approximation

• URL: http://arxiv.org/abs/2405.17834v2
• Date: Mon, 3 Jun 2024 14:05:52 GMT
• Title: Revisiting Step-Size Assumptions in Stochastic Approximation
• Authors: Caio Kalil Lauand, Sean Meyn,
• Abstract summary: The paper revisits step-size selection in a general Markovian setting. A major conclusion is that the choice of $rho =0$ or even $rho1/2$ is justified only in select settings.
• Score: 1.3654846342364308
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Many machine learning and optimization algorithms are built upon the framework of stochastic approximation (SA), for which the selection of step-size (or learning rate) is essential for success. For the sake of clarity, this paper focuses on the special case $\alpha_n = \alpha_0 n^{-\rho}$ at iteration $n$, with $\rho \in [0,1]$ and $\alpha_0>0$ design parameters. It is most common in practice to take $\rho=0$ (constant step-size), while in more theoretically oriented papers a vanishing step-size is preferred. In particular, with $\rho \in (1/2, 1)$ it is known that on applying the averaging technique of Polyak and Ruppert, the mean-squared error (MSE) converges at the optimal rate of $O(1/n)$ and the covariance in the central limit theorem (CLT) is minimal in a precise sense. The paper revisits step-size selection in a general Markovian setting. Under readily verifiable assumptions, the following conclusions are obtained provided $0<\rho<1$: $\bullet$ Parameter estimates converge with probability one, and also in $L_p$ for any $p\ge 1$. $\bullet$ The MSE may converge very slowly for small $\rho$, of order $O(\alpha_n^2)$ even with averaging. $\bullet$ For linear stochastic approximation the source of slow convergence is identified: for any $\rho\in (0,1)$, averaging results in estimates for which the error $\textit{covariance}$ vanishes at the optimal rate, and moreover the CLT covariance is optimal in the sense of Polyak and Ruppert. However, necessary and sufficient conditions are obtained under which the $\textit{bias}$ converges to zero at rate $O(\alpha_n)$. This is the first paper to obtain such strong conclusions while allowing for $\rho \le 1/2$. A major conclusion is that the choice of $\rho =0$ or even $\rho<1/2$ is justified only in select settings -- In general, bias may preclude fast convergence.

Related papers

• Optimal Sketching for Residual Error Estimation for Matrix and Vector Norms [50.15964512954274]
  We study the problem of residual error estimation for matrix and vector norms using a linear sketch. We demonstrate that this gives a substantial advantage empirically, for roughly the same sketch size and accuracy as in previous work. We also show an $Omega(k2/pn1-2/p)$ lower bound for the sparse recovery problem, which is tight up to a $mathrmpoly(log n)$ factor.
  arXiv  Detail & Related papers  (2024-08-16T02:33:07Z)
• Robust Distribution Learning with Local and Global Adversarial Corruptions [17.22168727622332]
  We develop an efficient finite-sample algorithm with error bounded by $sqrtvarepsilon k + rho + tildeO(dsqrtkn-1/(k lor 2))$ when $P$ has bounded covariance. Our efficient procedure relies on a novel trace norm approximation of an ideal yet intractable 2-Wasserstein projection estimator.
  arXiv  Detail & Related papers  (2024-06-10T17:48:36Z)
• On the $O(\frac{\sqrt{d}}{T^{1/4}})$ Convergence Rate of RMSProp and Its Momentum Extension Measured by $\ell_1$ Norm [59.65871549878937]
  This paper considers the RMSProp and its momentum extension and establishes the convergence rate of $frac1Tsum_k=1T. Our convergence rate matches the lower bound with respect to all the coefficients except the dimension $d$. Our convergence rate can be considered to be analogous to the $frac1Tsum_k=1T.
  arXiv  Detail & Related papers  (2024-02-01T07:21:32Z)
• High-dimensional Location Estimation via Norm Concentration for Subgamma Vectors [15.802475232604667]
  In location estimation, we are given $n$ samples from a known distribution $f$ shifted by an unknown translation $lambda$. Asymptotically, the maximum likelihood estimate achieves the Cram'er-Rao bound of error $mathcal N(0, frac1nmathcal I)$. We build on the theory using emphsmoothed estimators to bound the error for finite $n$ in terms of $mathcal I_r$, the Fisher information of the $r$-smoothed
  arXiv  Detail & Related papers  (2023-02-05T22:17:04Z)
• A spectral least-squares-type method for heavy-tailed corrupted regression with unknown covariance \& heterogeneous noise [2.019622939313173]
  We revisit heavy-tailed corrupted least-squares linear regression assuming to have a corrupted $n$-sized label-feature sample of at most $epsilon n$ arbitrary outliers. We propose a near-optimal computationally tractable estimator, based on the power method, assuming no knowledge on $(Sigma,Xi) nor the operator norm of $Xi$.
  arXiv  Detail & Related papers  (2022-09-06T23:37:31Z)
• Finite-Sample Maximum Likelihood Estimation of Location [16.44999338864628]
  For fixed $f$ the maximum-likelihood estimate (MLE) is well-known to be optimal in the limit as $n to infty$ We show for arbitrary $f$ and $n$ that one can recover a similar theory based on the Fisher information of a smoothed version of $f$, where the smoothing radius decays with $n$.
  arXiv  Detail & Related papers  (2022-06-06T04:33:41Z)
• TURF: A Two-factor, Universal, Robust, Fast Distribution Learning Algorithm [64.13217062232874]
  One of its most powerful and successful modalities approximates every distribution to an $ell$ distance essentially at most a constant times larger than its closest $t$-piece degree-$d_$. We provide a method that estimates this number near-optimally, hence helps approach the best possible approximation.
  arXiv  Detail & Related papers  (2022-02-15T03:49:28Z)
• Sparse sketches with small inversion bias [79.77110958547695]
  Inversion bias arises when averaging estimates of quantities that depend on the inverse covariance. We develop a framework for analyzing inversion bias, based on our proposed concept of an $(epsilon,delta)$-unbiased estimator for random matrices. We show that when the sketching matrix $S$ is dense and has i.i.d. sub-gaussian entries, the estimator $(epsilon,delta)$-unbiased for $(Atop A)-1$ with a sketch of size $m=O(d+sqrt d/
  arXiv  Detail & Related papers  (2020-11-21T01:33:15Z)
• Accelerating Optimization and Reinforcement Learning with Quasi-Stochastic Approximation [2.294014185517203]
  This paper sets out to extend convergence theory to quasi-stochastic approximations. It is illustrated with applications to gradient-free optimization and policy gradient algorithms for reinforcement learning.
  arXiv  Detail & Related papers  (2020-09-30T04:44:45Z)
• Model-Free Reinforcement Learning: from Clipped Pseudo-Regret to Sample Complexity [59.34067736545355]
  Given an MDP with $S$ states, $A$ actions, the discount factor $gamma in (0,1)$, and an approximation threshold $epsilon > 0$, we provide a model-free algorithm to learn an $epsilon$-optimal policy. For small enough $epsilon$, we show an improved algorithm with sample complexity.
  arXiv  Detail & Related papers  (2020-06-06T13:34:41Z)
• Agnostic Q-learning with Function Approximation in Deterministic Systems: Tight Bounds on Approximation Error and Sample Complexity [94.37110094442136]
  We study the problem of agnostic $Q$-learning with function approximation in deterministic systems. We show that if $delta = Oleft(rho/sqrtdim_Eright)$, then one can find the optimal policy using $Oleft(dim_Eright)$.
  arXiv  Detail & Related papers  (2020-02-17T18:41:49Z)

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.