arXiv subjects

Mert Gürbüzbalaban

Publications and source records attributed to Mert Gürbüzbalaban.

At least 19 recordsLinked to original sources

arXiv2026

RESIST: Resilient Decentralized Learning Using Consensus Gradient Descent

Empirical risk minimization (ERM) is a cornerstone of modern machine learning. This paper focuses on the man-in-the-middle (MITM) attack, wherein adversaries exploit communication vulnerabilities between devices to inject malicious updates during training. To address this challenge, we propose RESIST (Resilient dEcentralized learning using conSensus gradIent deScenT), an optimization algorithm designed to be robust against adversarially compromised communication links. RESIST uses a multistep consensus gradient descent framework with robust-statistics-based screening of neighbor messages. It has a design parameter $J$ that controls the frequency of local gradient computation: each local gradient update is preceded by $J-1$ communication and robust aggregation rounds. Compared with methods that perform both communication and a local gradient update at every iteration, RESIST with $J>2$ uses fewer local gradient updates at the same communication budget. We establish geometric algorithmic convergence guarantees for strongly convex and Polyak-Lojasiewicz ERM problems, and sublinear and finite-horizon guarantees for smooth nonconvex ERM problems. For heterogeneous local objectives, these guarantees quantify neighborhoods for the relevant iterate, objective-value, and stationarity errors. In the strongly convex homogeneous case, convergence becomes exact. We also establish statistical learning-rate guarantees under common-population independent and identically distributed sampling assumptions, including regimes in which the statistical error vanishes as the sample size grows. Experimental results demonstrate the robustness of RESIST across diverse attack strategies, screening methods, and loss functions. The $J$-ablation experiments further show that larger $J$ can improve convergence and reduce the number of local gradient updates at a fixed communication budget.

cs.LG↗

arXiv2026

A Stochastic GDA Method With Backtracking For Solving Nonconvex Concave Minimax Problems

We propose a stochastic GDA (gradient descent ascent) method with backtracking (SGDA-B) to solve nonconvex-concave (NCC) minimax problems of the form: $\min_{\mathbf{x}} \max_y \sum_{i=1}^N g_i(x_i)+f(\mathbf{x},y)-h(y)$, where $h$ and $g_i$ for $i=1,\cdots,N$ are closed, convex functions, and for some $L,μ\geq 0$, $f$ is $L$-smooth and $f(\mathbf{x},\cdot)$ is $μ$-strongly concave for all $\mathbf{x}$ in the problem domain. We consider the stochastic setting where one only has an access to an unbiased stochastic oracle of $\nabla f$ with a finite variance bound $σ^2$. While most of the existing methods assume knowledge of $L$, $μ$ and/or $σ^2$, SGDA-B is agnostic to all of these problem parameters. Moreover, SGDA-B can support random block-coordinate updates. In the deterministic setting, i.e., $σ^2=0$ and one can compute $\nabla f$ exactly, SGDA-B can compute an $ε$-stationary point within $\mathcal{O}(Lκ^2/ε^2)$ and $\mathcal{O}(L^3/ε^4)$ gradient calls when $μ>0$ and $μ=0$, respectively, where $κ\triangleq L/μ$. In the stochastic setting, i.e., $σ^2>0$, for any $p\in(0,1)$ and $ε>0$, it can compute an $ε$-stationary point with high probability, which requires $\mathcal{O}(L κ^3 ε^{-4} \log^2(1/p))$ and $\tilde{\mathcal{O}}(L^4ε^{-7}\log^2(1/p))$ stochastic oracle calls, with probability at least $1-p$, when $μ>0$ and $μ=0$, respectively. To our knowledge, SGDA-B is the first GDA-type method with backtracking to solve NCC minimax problems and achieves the best complexity among the methods that are agnostic to $L$, $μ$ and $σ^2$. We also provide numerical results for SGDA-B on a distributionally robust learning problem illustrating the potential performance gains that can be achieved by SGDA-B.

math.OC↗

arXiv2026

An Accelerated Primal Dual Algorithm with Backtracking for Decentralized Constrained Optimization

We propose a distributed accelerated primal-dual method with backtracking (D-APDB) for cooperative multi-agent constrained consensus optimization problems over an undirected network of agents, where only those agents connected by an edge can directly communicate to exchange large-volume data vectors using a high-speed, short-range communication protocol, e.g., WiFi, and we also assume that the network allows for one-hop simple information exchange beyond immediate neighbors as in LoRaWAN protocol. The objective is to minimize the sum of agent-specific composite convex functions over agent-specific private constraint sets. Unlike existing decentralized primal-dual methods that require knowledge of the Lipschitz constants, D-APDB automatically adapts to unknown smoothness constants by employing a distributed backtracking step-size search. Each agent relies only on first-order oracles associated with its own objective and constraint functions and on local communications with the neighboring agents, without any prior knowledge of Lipschitz constants. We establish $\mathcal{O}(1/K)$ convergence guarantees for sub-optimality, infeasibility and consensus violation, under standard assumptions on smoothness and on the connectivity of the communication graph. To our knowledge, when nodes have private constraints, especially when they are nonlinear convex constraints onto which projections are not cheap to compute, D-APDB is the first distributed method with backtracking that achieves the optimal convergence rate for the class of constrained composite convex optimization problems. We also provide numerical results for D-APDB on a distributed QCQP problem and distributed primal SVM training to illustrate the potential performance gains that can be achieved by D-APDB.

math.OC↗

arXiv2026

Accelerated Gradient Methods with Biased Gradient Estimates: Risk Sensitivity, High-Probability Guarantees, and Large Deviation Bounds

We study trade-offs between convergence rate and robustness to gradient errors in the context of first-order methods. Our focus is on generalized momentum methods (GMMs)--a broad class that includes Nesterov's accelerated gradient, heavy-ball, and gradient descent methods--for minimizing smooth strongly convex objectives. We allow stochastic gradient errors that may be adversarial and biased, and quantify robustness of these methods to gradient errors via the risk-sensitive index (RSI) from robust control theory. For quadratic objectives with i.i.d. Gaussian noise, we give closed form expressions for RSI in terms of solutions to 2x2 matrix Riccati equations, revealing a Pareto frontier between RSI and convergence rate over the choice of step-size and momentum parameters. We then prove a large-deviation principle for time-averaged suboptimality in the large iteration limit and show that the rate function is, up to a scaling, the convex conjugate of the RSI function. We further show that the rate function and RSI are linked to the $H_\infty$-norm--a measure of robustness to the worst-case deterministic gradient errors--so that stronger worst-case robustness (smaller $H_\infty$-norm) leads to sharper decay of the tail probabilities for the average suboptimality. Beyond quadratics, under potentially biased sub-Gaussian gradient errors, we derive non-asymptotic bounds on a finite-time analogue of the RSI, yielding finite-time high-probability guarantees and non-asymptotic large-deviation bounds for the averaged iterates. In the case of smooth strongly convex functions, we also observe an analogous trade-off between RSI and convergence-rate bounds. To our knowledge, these are the first non-asymptotic guarantees for GMMs with biased gradients and the first risk-sensitive analysis of GMMs. Finally, we provide numerical experiments on a robust regression problem to illustrate our results.

math.OC↗

arXiv2025

Rényi Differential Privacy for Heavy-Tailed SDEs via Fractional Poincaré Inequalities

Characterizing the differential privacy (DP) of learning algorithms has become a major challenge in recent years. In parallel, many studies suggested investigating the behavior of stochastic gradient descent (SGD) with heavy-tailed noise, both as a model for modern deep learning models and to improve their performance. However, most DP bounds focus on light-tailed noise, where satisfactory guarantees have been obtained but the proposed techniques do not directly extend to the heavy-tailed setting. Recently, the first DP guarantees for heavy-tailed SGD were obtained. These results provide $(0,δ)$-DP guarantees without requiring gradient clipping. Despite casting new light on the link between DP and heavy-tailed algorithms, these results have a strong dependence on the number of parameters and cannot be extended to other DP notions like the well-established Rényi differential privacy (RDP). In this work, we propose to address these limitations by deriving the first RDP guarantees for heavy-tailed SDEs, as well as their discretized counterparts. Our framework is based on new Rényi flow computations and the use of well-established fractional Poincaré inequalities. Under the assumption that such inequalities are satisfied, we obtain DP guarantees that have a much weaker dependence on the dimension compared to prior art.

stat.ML↗

arXiv2025

Privacy of SGD under Gaussian or Heavy-Tailed Noise: Guarantees without Gradient Clipping

The injection of heavy-tailed noise into the iterates of stochastic gradient descent (SGD) has garnered growing interest in recent years due to its theoretical and empirical benefits for optimization and generalization. However, its implications for privacy preservation remain largely unexplored. Aiming to bridge this gap, we provide differential privacy (DP) guarantees for noisy SGD, when the injected noise follows an $α$-stable distribution, which includes a spectrum of heavy-tailed distributions (with infinite variance) as well as the light-tailed Gaussian distribution. Considering the $(ε, δ)$-DP framework, we show that SGD with heavy-tailed perturbations achieves $(0, O(1/n))$-DP for a broad class of loss functions which can be non-convex, where $n$ is the number of data points. As a remarkable byproduct, contrary to prior work that necessitates bounded sensitivity for the gradients or clipping the iterates, our theory can handle unbounded gradients without clipping, and reveals that under mild assumptions, such a projection step is not actually necessary. Our results suggest that, given other benefits of heavy-tails in optimization, heavy-tailed noising schemes can be a viable alternative to their light-tailed counterparts.

stat.ML↗

arXiv2025

Entropic Risk-Averse Generalized Momentum Methods

In the context of first-order algorithms subject to random gradient noise, we study the trade-offs between the convergence rate (which quantifies how fast the initial conditions are forgotten) and the "risk" of suboptimality, i.e. deviations from the expected suboptimality. We focus on a general class of momentum methods (GMM) which recover popular methods such as gradient descent (GD), accelerated gradient descent (AGD), and heavy-ball (HB) method as special cases depending on the choice of GMM parameters. We use well-known risk measures "entropic risk" and "entropic value at risk" to quantify the risk of suboptimality. For strongly convex smooth minimization, we first obtain new convergence rate results for GMM with a unified theory that is also applicable to both AGD and HB, improving some of the existing results for HB. We then provide explicit bounds on the entropic risk and entropic value at risk of suboptimality at a given iterate which also provides direct bounds on the probability that the suboptimality exceeds a given threshold based on Chernoff's inequality. Our results unveil fundamental trade-offs between the convergence rate and the risk of suboptimality. We then plug the entropic risk and convergence rate estimates we obtained in a computationally tractable optimization framework and propose entropic risk-averse GMM (RA-GMM) and entropic risk-averse AGD (RA-AGD) methods which can select the GMM parameters to systematically trade-off the entropic value at risk with the convergence rate. We show that RA-AGD and RA-GMM lead to improved performance on quadratic optimization and logistic regression problems compared to the standard choice of parameters. To our knowledge, our work is the first to resort to coherent measures to design the parameters of momentum methods in a systematic manner.

math.OC↗

arXiv2024

High-probability complexity guarantees for nonconvex minimax problems

Stochastic smooth nonconvex minimax problems are prevalent in machine learning, e.g., GAN training, fair classification, and distributionally robust learning. Stochastic gradient descent ascent (GDA)-type methods are popular in practice due to their simplicity and single-loop nature. However, there is a significant gap between the theory and practice regarding high-probability complexity guarantees for these methods on stochastic nonconvex minimax problems. Existing high-probability bounds for GDA-type single-loop methods only apply to convex/concave minimax problems and to particular non-monotone variational inequality problems under some restrictive assumptions. In this work, we address this gap by providing the first high-probability complexity guarantees for nonconvex/PL minimax problems corresponding to a smooth function that satisfies the PL-condition in the dual variable. Specifically, we show that when the stochastic gradients are light-tailed, the smoothed alternating GDA method can compute an $\varepsilon$-stationary point within $O(\frac{\ell κ^2 δ^2}{\varepsilon^4} + \fracκ{\varepsilon^2}(\ell+δ^2\log({1}/{\bar{q}})))$ stochastic gradient calls with probability at least $1-\bar{q}$ for any $\bar{q}\in(0,1)$, where $μ$ is the PL constant, $\ell$ is the Lipschitz constant of the gradient, $κ=\ell/μ$ is the condition number, and $δ^2$ denotes a bound on the variance of stochastic gradients. We also present numerical results on a nonconvex/PL problem with synthetic data and on distributionally robust optimization problems with real data, illustrating our theoretical findings.

math.OC↗

arXiv2024

SAPD+: An Accelerated Stochastic Method for Nonconvex-Concave Minimax Problems

We propose a new stochastic method SAPD+ for solving nonconvex-concave minimax problems of the form $\min\max\mathcal{L}(x,y)=f(x)+Φ(x,y)-g(y)$, where $f,g$ are closed convex and $Φ(x,y)$ is a smooth function that is weakly convex in $x$, (strongly) concave in $y$. Let $δ^2$ denote the variance bound for the unbiased stochastic oracle used within SAPD+ to estimate $\nablaΦ$. When $δ>0$, for both strongly concave and merely concave settings, SAPD+ achieves the best known oracle complexities: $\mathcal{O}\Big(κ_y\max\Big\{1,\frac{δ^2}{ε^2}\Big\}\frac{L\mathcal{G}_0}{ε^{2}}\Big)$ for the strongly concave case without assuming compactness of the problem domain, and $\mathcal{O}\Big(\frac{L^3\mathcal{D}_y^2\mathcal{G}_0}{ε^{4}}\Big(1+\frac{δ^2}{ε^2}\Big)\Big)$ for the merely concave case, where $κ_y\geq 1$ is the condition number, $L$ is the Lipschitz constant of $\nabla Φ$, $\mathcal{G}_0$ is the primal-dual gap of the initial point, and $\mathcal{D}_y=\sup\{\|y\|:\ y\in\mathbf{dom} g\}$. We also propose SAPD+ with variance reduction, which enjoys $\mathcal{O}\Big(\max\Big\{κ_y,\sqrt{\fracδε}\Big\}\cdot (1+κ_y\fracδε)\frac{L\mathcal{G}_0}{ε^2}\Big)$ oracle complexity for weakly convex-strongly concave setting --this is the best known upper complexity bound in the literature for this setting and our paper establishes it for the first time. We demonstrate the efficiency of SAPD+ on a distributionally robust learning problem with a nonconvex regularizer and also on a multi-class classification problem in deep learning.

math.OC↗

arXiv2024

Robust Accelerated Primal-Dual Methods for Computing Saddle Points

We consider strongly-convex-strongly-concave saddle point problems assuming we have access to unbiased stochastic estimates of the gradients. We propose a stochastic accelerated primal-dual (SAPD) algorithm and show that SAPD sequence, generated using constant primal-dual step sizes, linearly converges to a neighborhood of the unique saddle point. Interpreting the size of the neighborhood as a measure of robustness to gradient noise, we obtain explicit characterizations of robustness in terms of SAPD parameters and problem constants. Based on these characterizations, we develop computationally tractable techniques for optimizing the SAPD parameters, i.e., the primal and dual step sizes, and the momentum parameter, to achieve a desired trade-off between the convergence rate and robustness on the Pareto curve. This allows SAPD to enjoy fast convergence properties while being robust to noise as an accelerated method. SAPD admits convergence guarantees for the distance metric with a variance term optimal up to a logarithmic factor -which can be removed by employing a restarting strategy. We also discuss how convergence and robustness results extend to the convex-concave setting. Finally, we illustrate our framework on distributionally robust logistic regression problem.

math.OC↗

arXiv2024

Penalized Overdamped and Underdamped Langevin Monte Carlo Algorithms for Constrained Sampling

We consider the constrained sampling problem where the goal is to sample from a target distribution $π(x)\propto e^{-f(x)}$ when $x$ is constrained to lie on a convex body $\mathcal{C}$. Motivated by penalty methods from continuous optimization, we propose penalized Langevin Dynamics (PLD) and penalized underdamped Langevin Monte Carlo (PULMC) methods that convert the constrained sampling problem into an unconstrained sampling problem by introducing a penalty function for constraint violations. When $f$ is smooth and gradients are available, we get $\tilde{\mathcal{O}}(d/\varepsilon^{10})$ iteration complexity for PLD to sample the target up to an $\varepsilon$-error where the error is measured in the TV distance and $\tilde{\mathcal{O}}(\cdot)$ hides logarithmic factors. For PULMC, we improve the result to $\tilde{\mathcal{O}}(\sqrt{d}/\varepsilon^{7})$ when the Hessian of $f$ is Lipschitz and the boundary of $\mathcal{C}$ is sufficiently smooth. To our knowledge, these are the first convergence results for underdamped Langevin Monte Carlo methods in the constrained sampling that handle non-convex $f$ and provide guarantees with the best dimension dependency among existing methods with deterministic gradient. If unbiased stochastic estimates of the gradient of $f$ are available, we propose PSGLD and PSGULMC methods that can handle stochastic gradients and are scaleable to large datasets without requiring Metropolis-Hasting correction steps. For PSGLD and PSGULMC, when $f$ is strongly convex and smooth, we obtain $\tilde{\mathcal{O}}(d/\varepsilon^{18})$ and $\tilde{\mathcal{O}}(d\sqrt{d}/\varepsilon^{39})$ iteration complexity in W2 distance. When $f$ is smooth and can be non-convex, we provide finite-time performance bounds and iteration complexity results. Finally, we illustrate the performance on Bayesian LASSO regression and Bayesian constrained deep learning problems.

stat.ML↗

arXiv2023

Cyclic and Randomized Stepsizes Invoke Heavier Tails in SGD than Constant Stepsize

Cyclic and randomized stepsizes are widely used in the deep learning practice and can often outperform standard stepsize choices such as constant stepsize in SGD. Despite their empirical success, not much is currently known about when and why they can theoretically improve the generalization performance. We consider a general class of Markovian stepsizes for learning, which contain i.i.d. random stepsize, cyclic stepsize as well as the constant stepsize as special cases, and motivated by the literature which shows that heaviness of the tails (measured by the so-called "tail-index") in the SGD iterates is correlated with generalization, we study tail-index and provide a number of theoretical results that demonstrate how the tail-index varies on the stepsize scheduling. Our results bring a new understanding of the benefits of cyclic and randomized stepsizes compared to constant stepsize in terms of the tail behavior. We illustrate our theory on linear regression experiments and show through deep learning experiments that Markovian stepsizes can achieve even a heavier tail and be a viable alternative to cyclic and i.i.d. randomized stepsize rules.

stat.ML↗

arXiv2023

High Probability and Risk-Averse Guarantees for a Stochastic Accelerated Primal-Dual Method

We consider stochastic strongly-convex-strongly-concave (SCSC) saddle point (SP) problems which frequently arise in applications ranging from distributionally robust learning to game theory and fairness in machine learning. We focus on the recently developed stochastic accelerated primal-dual algorithm (SAPD), which admits optimal complexity in several settings as an accelerated algorithm. We provide high probability guarantees for convergence to a neighborhood of the saddle point that reflects accelerated convergence behavior. We also provide an analytical formula for the limiting covariance matrix of the iterates for a class of stochastic SCSC quadratic problems where the gradient noise is additive and Gaussian. This allows us to develop lower bounds for this class of quadratic problems which show that our analysis is tight in terms of the high probability bound dependency to the parameters. We also provide a risk-averse convergence analysis characterizing the ``Conditional Value at Risk'', the ``Entropic Value at Risk'', and the $χ^2$-divergence of the distance to the saddle point, highlighting the trade-offs between the bias and the risk associated with an approximate solution obtained by terminating the algorithm at any iteration.

math.OC↗

arXiv2023

Distributionally Robust Learning with Weakly Convex Losses: Convergence Rates and Finite-Sample Guarantees

We consider a distributionally robust stochastic optimization problem and formulate it as a stochastic two-level composition optimization problem with the use of the mean--semideviation risk measure. In this setting, we consider a single time-scale algorithm, involving two versions of the inner function value tracking: linearized tracking of a continuously differentiable loss function, and SPIDER tracking of a weakly convex loss function. We adopt the norm of the gradient of the Moreau envelope as our measure of stationarity and show that the sample complexity of $\mathcal{O}(\varepsilon^{-3})$ is possible in both cases, with only the constant larger in the second case. Finally, we demonstrate the performance of our algorithm with a robust learning example and a weakly convex, non-smooth regression example.

math.OC↗

arXiv2023

Algorithmic Stability of Heavy-Tailed Stochastic Gradient Descent on Least Squares

Recent studies have shown that heavy tails can emerge in stochastic optimization and that the heaviness of the tails have links to the generalization error. While these studies have shed light on interesting aspects of the generalization behavior in modern settings, they relied on strong topological and statistical regularity assumptions, which are hard to verify in practice. Furthermore, it has been empirically illustrated that the relation between heavy tails and generalization might not always be monotonic in practice, contrary to the conclusions of existing theory. In this study, we establish novel links between the tail behavior and generalization properties of stochastic gradient descent (SGD), through the lens of algorithmic stability. We consider a quadratic optimization problem and use a heavy-tailed stochastic differential equation (and its Euler discretization) as a proxy for modeling the heavy-tailed behavior emerging in SGD. We then prove uniform stability bounds, which reveal the following outcomes: (i) Without making any exotic assumptions, we show that SGD will not be stable if the stability is measured with the squared-loss $x\mapsto x^2$, whereas it in turn becomes stable if the stability is instead measured with a surrogate loss $x\mapsto |x|^p$ with some $p<2$. (ii) Depending on the variance of the data, there exists a \emph{`threshold of heavy-tailedness'} such that the generalization error decreases as the tails become heavier, as long as the tails are lighter than this threshold. This suggests that the relation between heavy tails and generalization is not globally monotonic. (iii) We prove matching lower-bounds on uniform stability, implying that our bounds are tight in terms of the heaviness of the tails. We support our theory with synthetic and real neural network experiments.

stat.ML↗

arXiv2023

Algorithmic Stability of Heavy-Tailed SGD with General Loss Functions

Heavy-tail phenomena in stochastic gradient descent (SGD) have been reported in several empirical studies. Experimental evidence in previous works suggests a strong interplay between the heaviness of the tails and generalization behavior of SGD. To address this empirical phenomena theoretically, several works have made strong topological and statistical assumptions to link the generalization error to heavy tails. Very recently, new generalization bounds have been proven, indicating a non-monotonic relationship between the generalization error and heavy tails, which is more pertinent to the reported empirical observations. While these bounds do not require additional topological assumptions given that SGD can be modeled using a heavy-tailed stochastic differential equation (SDE), they can only apply to simple quadratic problems. In this paper, we build on this line of research and develop generalization bounds for a more general class of objective functions, which includes non-convex functions as well. Our approach is based on developing Wasserstein stability bounds for heavy-tailed SDEs and their discretizations, which we then convert to generalization bounds. Our results do not require any nontrivial assumptions; yet, they shed more light to the empirical observations, thanks to the generality of the loss functions.

stat.ML↗

arXiv2022

TENGraD: Time-Efficient Natural Gradient Descent with Exact Fisher-Block Inversion

This work proposes a time-efficient Natural Gradient Descent method, called TENGraD, with linear convergence guarantees. Computing the inverse of the neural network's Fisher information matrix is expensive in NGD because the Fisher matrix is large. Approximate NGD methods such as KFAC attempt to improve NGD's running time and practical application by reducing the Fisher matrix inversion cost with approximation. However, the approximations do not reduce the overall time significantly and lead to less accurate parameter updates and loss of curvature information. TENGraD improves the time efficiency of NGD by computing Fisher block inverses with a computationally efficient covariance factorization and reuse method. It computes the inverse of each block exactly using the Woodbury matrix identity to preserve curvature information while admitting (linear) fast convergence rates. Our experiments on image classification tasks for state-of-the-art deep neural architecture on CIFAR-10, CIFAR-100, and Fashion-MNIST show that TENGraD significantly outperforms state-of-the-art NGD methods and often stochastic gradient descent in wall-clock time.

cs.LG↗

arXiv2022

Why Random Reshuffling Beats Stochastic Gradient Descent

We analyze the convergence rate of the random reshuffling (RR) method, which is a randomized first-order incremental algorithm for minimizing a finite sum of convex component functions. RR proceeds in cycles, picking a uniformly random order (permutation) and processing the component functions one at a time according to this order, i.e., at each cycle, each component function is sampled without replacement from the collection. Though RR has been numerically observed to outperform its with-replacement counterpart stochastic gradient descent (SGD), characterization of its convergence rate has been a long standing open question. In this paper, we answer this question by showing that when the component functions are quadratics or smooth and the sum function is strongly convex, RR with iterate averaging and a diminishing stepsize $α_k=Θ(1/k^s)$ for $s\in (1/2,1)$ converges at rate $Θ(1/k^{2s})$ with probability one in the suboptimality of the objective value, thus improving upon the $Ω(1/k)$ rate of SGD. Our analysis draws on the theory of Polyak-Ruppert averaging and relies on decoupling the dependent cycle gradient error into an independent term over cycles and another term dominated by $α_k^2$. This allows us to apply law of large numbers to an appropriately weighted version of the cycle gradient errors, where the weights depend on the stepsize. We also provide high probability convergence rate estimates that shows decay rate of different terms and allows us to propose a modification of RR with convergence rate ${\cal O}(\frac{1}{k^2})$.

math.OC↗