Arrow Research search

Author name cluster

Patrick Rebeschini

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

28 papers
2 author rows

Possible papers

28

EWRL Workshop 2025 Workshop Paper

Does Stochastic Gradient really succeed for bandits?

  • Dorian Baudry
  • Emmeran Johnson
  • Simon Vary
  • Ciara Pike-Burke
  • Patrick Rebeschini

Recent works of Mei et al. (2023, 2024) have deepened the theoretical understanding of the Stochastic Gradient Bandit (SGB) policy, showing that using a constant learning rate guarantees asymptotic convergence to the optimal policy, and that sufficiently small learning rates can yield logarithmic regret. However, whether logarithmic regret holds beyond small learning rates remains unclear. In this work, we take a step towards characterizing the regret regimes of SGB as a function of its learning rate. For two--armed bandits, we identify a sharp threshold, scaling with the sub-optimality gap $\Delta$, below which SGB achieves logarithmic regret on all instances, and above which it can incur polynomial regret on some instances. This result highlights the necessity of knowing (or estimating) $\Delta$ to ensure logarithmic regret with a constant learning rate. For general $K$-armed bandits, we further show the learning rate must scale inversely with $K$ to avoid polynomial regret. We introduce novel techniques to derive regret upper bounds for SGB, laying the groundwork for future advances in the theory of gradient-based bandit algorithms.

PDF

NeurIPS Conference 2025 Conference Paper

Does Stochastic Gradient really succeed for bandits?

  • Dorian Baudry
  • Emmeran Johnson
  • Simon Vary
  • Ciara Pike-Burke
  • Patrick Rebeschini

Recent works of Mei et al. (2023, 2024) have deepened the theoretical understanding of the *Stochastic Gradient Bandit* (SGB) policy, showing that using a constant learning rate guarantees asymptotic convergence to the optimal policy, and that sufficiently *small* learning rates can yield logarithmic regret. However, whether logarithmic regret holds beyond small learning rates remains unclear. In this work, we take a step towards characterizing the regret *regimes* of SGB as a function of its learning rate. For two--armed bandits, we identify a sharp threshold, scaling with the sub-optimality gap $\Delta$, below which SGB achieves *logarithmic* regret on all instances, and above which it can incur *polynomial* regret on some instances. This result highlights the necessity of knowing (or estimating) $\Delta$ to ensure logarithmic regret with a constant learning rate. For general $K$-armed bandits, we further show the learning rate must scale inversely with $K$ to avoid polynomial regret. We introduce novel techniques to derive regret upper bounds for SGB, laying the groundwork for future advances in the theory of gradient-based bandit algorithms.

PDF Details

ICLR Conference 2025 Conference Paper

Learning mirror maps in policy mirror descent

  • Carlo Alfano
  • Sebastian Rene Towers
  • Silvia Sapora
  • Chris Lu 0001
  • Patrick Rebeschini

Policy Mirror Descent (PMD) is a popular framework in reinforcement learning, serving as a unifying perspective that encompasses numerous algorithms. These algorithms are derived through the selection of a mirror map and enjoy finite-time convergence guarantees. Despite its popularity, the exploration of PMD's full potential is limited, with the majority of research focusing on a particular mirror map---namely, the negative entropy---which gives rise to the renowned Natural Policy Gradient (NPG) method. It remains uncertain from existing theoretical studies whether the choice of mirror map significantly influences PMD's efficacy. In our work, we conduct empirical investigations to show that the conventional mirror map choice (NPG) often yields less-than-optimal outcomes across several standard benchmark environments. Using evolutionary strategies, we identify more efficient mirror maps that enhance the performance of PMD. We first focus on a tabular environment, i.e.\ Grid-World, where we relate existing theoretical bounds with the performance of PMD for a few standard mirror maps and the learned one. We then show that it is possible to learn a mirror map that outperforms the negative entropy in more complex environments, such as the MinAtar suite. Additionally, we demonstrate that the learned mirror maps generalize effectively to different tasks by testing each map across various other environments.

Details

NeurIPS Conference 2025 Conference Paper

Meta-Learning Objectives for Preference Optimization

  • Carlo Alfano
  • Silvia Sapora
  • Jakob Foerster
  • Patrick Rebeschini
  • Yee Whye Teh

Evaluating preference optimization (PO) algorithms on LLM alignment is a challenging task that presents prohibitive costs, noise, and several variables like model size and hyper-parameters. In this work, we show that it is possible to gain insights on the efficacy of PO algorithm on simpler benchmarks. We design a diagnostic suite of MuJoCo tasks and datasets, which we use to systematically evaluate PO algorithms, establishing a more controlled and cheaper benchmark. We then propose a novel family of PO algorithms based on mirror descent, which we call Mirror Preference Optimization (MPO). Through evolutionary strategies, we search this class to discover algorithms specialized to specific properties of preference datasets, such as mixed-quality or noisy data. We demonstrate that our discovered PO algorithms outperform all known algorithms in the targeted MuJoCo settings. Finally, based on the insights gained from our MuJoCo experiments, we design a PO algorithm that significantly outperform existing baselines in an LLM alignment task.

PDF Details

NeurIPS Conference 2025 Conference Paper

Non-stationary Bandit Convex Optimization: A Comprehensive Study

  • Xiaoqi Liu
  • Dorian Baudry
  • Julian Zimmert
  • Patrick Rebeschini
  • Arya Akhavan

Bandit Convex Optimization is a fundamental class of sequential decision-making problems, where the learner selects actions from a continuous domain and observes a loss (but not its gradient) at only one point per round. We study this problem in non-stationary environments, and aim to minimize the regret under three standard measures of non-stationarity: the number of switches $S$ in the comparator sequence, the total variation $\Delta$ of the loss functions, and the path-length $P$ of the comparator sequence. We propose a polynomial-time algorithm, Tilted Exponentially Weighted Average with Sleeping Experts (TEWA-SE), which adapts the sleeping experts framework from online convex optimization to the bandit setting. For strongly convex losses, we prove that TEWA-SE is minimax-optimal with respect to known $S$ and $\Delta$ by establishing matching upper and lower bounds. By equipping TEWA-SE with the Bandit-over-Bandit framework, we extend our analysis to environments with unknown non-stationarity measures. For general convex losses, we introduce a second algorithm, clipped Exploration by Optimization (cExO), based on exponential weights over a discretized action space. While not polynomial-time computable, this method achieves minimax-optimal regret with respect to known $S$ and $\Delta$, and improves on the best existing bounds with respect to $P$.

PDF Details

NeurIPS Conference 2025 Conference Paper

On the necessity of adaptive regularisation: Optimal anytime online learning on $\boldsymbol{\ell_p}$-balls

  • Emmeran Johnson
  • David Martínez-Rubio
  • Ciara Pike-Burke
  • Patrick Rebeschini

We study online convex optimization on $\ell_p$-balls in $\mathbb{R}^d$ for $p > 2$. While always sub-linear, the optimal regret exhibits a shift between the high-dimensional setting ($d > T$), when the dimension $d$ is greater than the time horizon $T$ and the low-dimensional setting ($d \leq T$). We show that Follow-the-Regularised-Leader (FTRL) with time-varying regularisation which is adaptive to the dimension regime is anytime optimal for all dimension regimes. Motivated by this, we ask whether it is possible to obtain anytime optimality of FTRL with fixed non-adaptive regularisation. Our main result establishes that for separable regularisers, adaptivity in the regulariser is necessary, and that any fixed regulariser will be sub-optimal in one of the two dimension regimes. Finally, we provide lower bounds which rule out sub-linear regret bounds for the linear bandit problem in sufficiently high-dimension for all $\ell_p$-balls with $p \geq 1$.

PDF Details

EWRL Workshop 2025 Workshop Paper

Stochastic Shortest Path with Sparse Adversarial Costs

  • Emmeran Johnson
  • Alberto Rumi
  • Ciara Pike-Burke
  • Patrick Rebeschini

We study the adversarial Stochastic Shortest Path (SSP) problem with sparse costs under full-information feedback. In the known transition setting, existing bounds based on Online Mirror Descent (OMD) with negative-entropy regularization scale with $\sqrt{\log S A}$, where $SA$ is the size of the state-action space. While we show that this is optimal in the worst-case, this bound fails to capture the benefits of sparsity when only a small number $M \ll SA$ of state-action pairs incur cost. In fact, we also show that the negative-entropy is inherently non-adaptive to sparsity: it provably incurs regret scaling with $\sqrt{\log S}$ on sparse problems. Instead, we propose a novel family of $\ell_r$-norm regularizers ($r \in (1, 2)$) that adapts to the sparsity and achieves regret scaling with $\sqrt{\log M}$ instead of $\sqrt{\log SA}$. We show this is optimal via a matching lower bound, highlighting that $M$ captures the effective dimension of the problem instead of $SA$. Finally, in the unknown transition setting the benefits of sparsity are limited: we prove that even on sparse problems, the minimax regret for any learner scales polynomially with $SA$.

PDF

NeurIPS Conference 2025 Conference Paper

Stochastic Shortest Path with Sparse Adversarial Costs

  • Emmeran Johnson
  • Alberto Rumi
  • Ciara Pike-Burke
  • Patrick Rebeschini

We study the adversarial Stochastic Shortest Path (SSP) problem with sparse costs under full-information feedback. In the known transition setting, existing bounds based on Online Mirror Descent (OMD) with negative-entropy regularization scale with $\sqrt{\log S A}$, where $SA$ is the size of the state-action space. While we show that this is optimal in the worst-case, this bound fails to capture the benefits of sparsity when only a small number $M \ll SA$ of state-action pairs incur cost. In fact, we also show that the negative-entropy is inherently non-adaptive to sparsity: it provably incurs regret scaling with $\sqrt{\log S}$ on sparse problems. Instead, we propose a novel family of $\ell_r$-norm regularizers ($r \in (1, 2)$) that adapts to the sparsity and achieves regret scaling with $\sqrt{\log M}$ instead of $\sqrt{\log SA}$. We show this is optimal via a matching lower bound, highlighting that $M$ captures the effective dimension of the problem instead of $SA$. Finally, in the unknown transition setting the benefits of sparsity are limited: we prove that even on sparse problems, the minimax regret for any learner scales polynomially with $SA$.

PDF Details

JMLR Journal 2024 Journal Article

Exponential Tail Local Rademacher Complexity Risk Bounds Without the Bernstein Condition

  • Varun Kanade
  • Patrick Rebeschini
  • Tomas Vaskevicius

The local Rademacher complexity framework is one of the most successful general-purpose toolboxes for establishing sharp excess risk bounds for statistical estimators based on empirical risk minimization. However, applying this toolbox typically requires using the Bernstein condition, which often restricts the applicability domain to convex and proper settings. Recent years have witnessed several examples of problems where optimal statistical performance is only achievable via non-convex and improper estimators originating from aggregation theory, including the fundamental problem of model selection. These examples are currently outside the reach of the classical local Rademacher complexity theory. In this work, we build upon the recent approach to localization via offset Rademacher complexities, for which a general high-probability theory has yet to be established. Our main result is an exponential-tail offset Rademacher complexity excess risk upper bound that yields results at least as sharp as those obtainable via the classical theory. However, our bound applies under an estimator-dependent geometric condition (the “offset condition”) instead of the estimator-independent (but, in general, distribution-dependent) Bernstein condition on which the classical theory relies. Our results apply to improper prediction regimes not directly covered by the classical theory, such as optimal model selection aggregation for arbitrary classes (including infinite and non-convex classes), and early-stopping/iterative regularization; the Bernstein condition does not hold in both examples. [abs] [ pdf ][ bib ] &copy JMLR 2024. ( edit, beta )

PDF Details

EWRL Workshop 2024 Workshop Paper

Learning mirror maps in policy mirror descent

  • Carlo Alfano
  • Sebastian Rene Towers
  • Silvia Sapora
  • Chris Lu
  • Patrick Rebeschini

Policy Mirror Descent (PMD) is a popular framework in reinforcement learning, serving as a unifying perspective that encompasses numerous algorithms. These algorithms are derived through the selection of a mirror map and enjoy finite-time convergence guarantees. Despite its popularity, the exploration of PMD's full potential is limited, with the majority of research focusing on a particular mirror map---namely, the negative entropy---which gives rise to the renowned Natural Policy Gradient (NPG) method. It remains uncertain from existing theoretical studies whether the choice of mirror map significantly influences PMD's efficacy. In our work, we conduct empirical investigations to show that the conventional mirror map choice (NPG) often yields less-than-optimal outcomes across several standard benchmark environments. Using evolutionary strategies, we identify more efficient mirror maps that enhance the performance of PMD. We first focus on a tabular environment, i. e. Grid-World, where we relate existing theoretical bounds with the performance of PMD for a few standard mirror maps and the learned one. We then show that it is possible to learn a mirror map that outperforms the negative entropy in more complex environments, such as the MinAtar suite. Our results suggest that mirror maps generalize well across various environments, raising questions about how to best match a mirror map to an environment's structure and characteristics.

PDF

ICLR Conference 2024 Conference Paper

Sample-Efficiency in Multi-Batch Reinforcement Learning: The Need for Dimension-Dependent Adaptivity

  • Emmeran Johnson
  • Ciara Pike-Burke
  • Patrick Rebeschini

We theoretically explore the relationship between sample-efficiency and adaptivity in reinforcement learning. An algorithm is sample-efficient if it uses a number of queries $n$ to the environment that is polynomial in the dimension $d$ of the problem. Adaptivity refers to the frequency at which queries are sent and feedback is processed to update the querying strategy. To investigate this interplay, we employ a learning framework that allows sending queries in $K$ batches, with feedback being processed and queries updated after each batch. This model encompasses the whole adaptivity spectrum, ranging from non-adaptive `offline' ($K=1$) to fully adaptive ($K=n$) scenarios, and regimes in between. For the problems of policy evaluation and best-policy identification under $d$-dimensional linear function approximation, we establish $\Omega(\log \log d)$ lower bounds on the number of batches $K$ required for sample-efficient algorithms with $n = O(poly(d))$ queries. Our results show that just having adaptivity ($K>1$) does not necessarily guarantee sample-efficiency. Notably, the adaptivity-boundary for sample-efficiency is not between offline reinforcement learning ($K=1$), where sample-efficiency was known to not be possible, and adaptive settings. Instead, the boundary lies between different regimes of adaptivity and depends on the problem dimension.

Details

NeurIPS Conference 2023 Conference Paper

A Novel Framework for Policy Mirror Descent with General Parameterization and Linear Convergence

  • Carlo Alfano
  • Rui Yuan
  • Patrick Rebeschini

Modern policy optimization methods in reinforcement learning, such as TRPO and PPO, owe their success to the use of parameterized policies. However, while theoretical guarantees have been established for this class of algorithms, especially in the tabular setting, the use of general parameterization schemes remains mostly unjustified. In this work, we introduce a novel framework for policy optimization based on mirror descent that naturally accommodates general parameterizations. The policy class induced by our scheme recovers known classes, e. g. , softmax, and generates new ones depending on the choice of mirror map. Using our framework, we obtain the first result that guarantees linear convergence for a policy-gradient-based method involving general parameterization. To demonstrate the ability of our framework to accommodate general parameterization schemes, we provide its sample complexity when using shallow neural networks, show that it represents an improvement upon the previous best results, and empirically validate the effectiveness of our theoretical claims on classic control tasks.

PDF Details

EWRL Workshop 2023 Workshop Paper

A Novel Framework for Policy Mirror Descent with General Parametrization and Linear Convergence

  • Carlo Alfano
  • Rui Yuan
  • Patrick Rebeschini

Modern policy optimization methods in reinforcement learning, such as Trust Region Policy Optimization and Proximal Policy Optimization, owe their success to the use of parameterized policies. However, while theoretical guarantees have been established for this class of algorithms, especially in the tabular setting, the use of general parametrization schemes remains mostly unjustified. In this work, we introduce a novel framework for policy optimization based on mirror descent that naturally accommodates general parametrizations. The policy class induced by our scheme recovers known classes, e. g. \ softmax, and generates new ones depending on the choice of mirror map. For our framework, we obtain the first result that guarantees linear convergence for a policy-gradient-based method involving general parametrization. To demonstrate the ability of our framework to accommodate general parametrization schemes, we obtain its sample complexity when using shallow neural networks and show that it represents an improvement upon the previous best results.

PDF

NeurIPS Conference 2023 Conference Paper

Optimal Convergence Rate for Exact Policy Mirror Descent in Discounted Markov Decision Processes

  • Emmeran Johnson
  • Ciara Pike-Burke
  • Patrick Rebeschini

Policy Mirror Descent (PMD) is a general family of algorithms that covers a wide range of novel and fundamental methods in reinforcement learning. Motivated by the instability of policy iteration (PI) with inexact policy evaluation, unregularised PMD algorithmically regularises the policy improvement step of PI without regularising the objective function. With exact policy evaluation, PI is known to converge linearly with a rate given by the discount factor $\gamma$ of a Markov Decision Process. In this work, we bridge the gap between PI and PMD with exact policy evaluation and show that the dimension-free $\gamma$-rate of PI can be achieved by the general family of unregularised PMD algorithms under an adaptive step-size. We show that both the rate and step-size are unimprovable for PMD: we provide matching lower bounds that demonstrate that the $\gamma$-rate is optimal for PMD methods as well as PI and that the adaptive step-size is necessary to achieve it. Our work is the first to relate PMD to rate-optimality and step-size necessity. Our study of the convergence of PMD avoids the use of the performance difference lemma, which leads to a direct analysis of independent interest. We also extend the analysis to the inexact setting and establish the first dimension-optimal sample complexity for unregularised PMD under a generative model, improving upon the best-known result.

PDF Details

EWRL Workshop 2023 Workshop Paper

Optimal Convergence Rate for Exact Policy Mirror Descent in Discounted Markov Decision Processes

  • Emmeran Johnson
  • Ciara Pike-Burke
  • Patrick Rebeschini

Policy Mirror Descent (PMD) is a general family of algorithms that covers a wide range of novel and fundamental methods in reinforcement learning. Motivated by the instability of policy iteration (PI) with inexact policy evaluation, PMD algorithmically regularises the policy improvement step of PI. With exact policy evaluation, PI is known to converge linearly with a rate given by the discount factor $\gamma$ of a Markov Decision Process. In this work, we bridge the gap between PI and PMD with exact policy evaluation and show that the dimension-free $\gamma$-rate of PI can be achieved by the general family of unregularised PMD algorithms under an adaptive step-size. We show that both the rate and step-size are unimprovable for PMD: we provide matching lower bounds that demonstrate that the $\gamma$-rate is optimal for PMD methods as well as PI and that the adaptive step-size is necessary to achieve it. Our work is the first to relate PMD to rate-optimality and step-size necessity. Our study of the convergence of PMD avoids the use of the performance difference lemma, which leads to a direct analysis of independent interest. We also extend the analysis to the inexact setting and establish the first dimension-optimal sample complexity for unregularised PMD under a generative model, improving upon the best-known result.

PDF

NeurIPS Conference 2021 Conference Paper

Distributed Machine Learning with Sparse Heterogeneous Data

  • Dominic Richards
  • Sahand Negahban
  • Patrick Rebeschini

Motivated by distributed machine learning settings such as Federated Learning, we consider the problem of fitting a statistical model across a distributed collection of heterogeneous data sets whose similarity structure is encoded by a graph topology. Precisely, we analyse the case where each node is associated with fitting a sparse linear model, and edges join two nodes if the difference of their solutions is also sparse. We propose a method based on Basis Pursuit Denoising with a total variation penalty, and provide finite sample guarantees for sub-Gaussian design matrices. Taking the root of the tree as a reference node, we show that if the sparsity of the differences across nodes is smaller than the sparsity at the root, then recovery is successful with fewer samples than by solving the problems independently, or by using methods that rely on a large overlap in the signal supports, such as the group Lasso. We consider both the noiseless and noisy setting, and numerically investigate the performance of distributed methods based on Distributed Alternating Direction Methods of Multipliers (ADMM) and hyperspectral unmixing.

PDF Details

NeurIPS Conference 2021 Conference Paper

Implicit Regularization in Matrix Sensing via Mirror Descent

  • Fan Wu
  • Patrick Rebeschini

We study discrete-time mirror descent applied to the unregularized empirical risk in matrix sensing. In both the general case of rectangular matrices and the particular case of positive semidefinite matrices, a simple potential-based analysis in terms of the Bregman divergence allows us to establish convergence of mirror descent---with different choices of the mirror maps---to a matrix that, among all global minimizers of the empirical risk, minimizes a quantity explicitly related to the nuclear norm, the Frobenius norm, and the von Neumann entropy. In both cases, this characterization implies that mirror descent, a first-order algorithm minimizing the unregularized empirical risk, recovers low-rank matrices under the same set of assumptions that are sufficient to guarantee recovery for nuclear-norm minimization. When the sensing matrices are symmetric and commute, we show that gradient descent with full-rank factorized parametrization is a first-order approximation to mirror descent, in which case we obtain an explicit characterization of the implicit bias of gradient flow as a by-product.

PDF Details

NeurIPS Conference 2021 Conference Paper

On Optimal Interpolation in Linear Regression

  • Eduard Oravkin
  • Patrick Rebeschini

Understanding when and why interpolating methods generalize well has recently been a topic of interest in statistical learning theory. However, systematically connecting interpolating methods to achievable notions of optimality has only received partial attention. In this paper, we ask the question of what is the optimal way to interpolate in linear regression using functions that are linear in the response variable (as the case for the Bayes optimal estimator in ridge regression) and depend on the data, the population covariance of the data, the signal-to-noise ratio and the covariance of the prior for the signal, but do not depend on the value of the signal itself nor the noise vector in the training data. We provide a closed-form expression for the interpolator that achieves this notion of optimality and show that it can be derived as the limit of preconditioned gradient descent with a specific initialization. We identify a regime where the minimum-norm interpolator provably generalizes arbitrarily worse than the optimal response-linear achievable interpolator that we introduce, and validate with numerical experiments that the notion of optimality we consider can be achieved by interpolating methods that only use the training data as input in the case of an isotropic prior. Finally, we extend the notion of optimal response-linear interpolation to random features regression under a linear data-generating model.

PDF Details

NeurIPS Conference 2021 Conference Paper

Time-independent Generalization Bounds for SGLD in Non-convex Settings

  • Tyler Farghly
  • Patrick Rebeschini

We establish generalization error bounds for stochastic gradient Langevin dynamics (SGLD) with constant learning rate under the assumptions of dissipativity and smoothness, a setting that has received increased attention in the sampling/optimization literature. Unlike existing bounds for SGLD in non-convex settings, ours are time-independent and decay to zero as the sample size increases. Using the framework of uniform stability, we establish time-independent bounds by exploiting the Wasserstein contraction property of the Langevin diffusion, which also allows us to circumvent the need to bound gradients using Lipschitz-like assumptions. Our analysis also supports variants of SGLD that use different discretization methods, incorporate Euclidean projections, or use non-isotropic noise.

PDF Details

NeurIPS Conference 2020 Conference Paper

A Continuous-Time Mirror Descent Approach to Sparse Phase Retrieval

  • Fan Wu
  • Patrick Rebeschini

We analyze continuous-time mirror descent applied to sparse phase retrieval, which is the problem of recovering sparse signals from a set of magnitude-only measurements. We apply mirror descent to the unconstrained empirical risk minimization problem (batch setting), using the square loss and square measurements. We provide a full convergence analysis of the algorithm in this non-convex setting and prove that, with the hypentropy mirror map, mirror descent recovers any $k$-sparse vector $\mathbf{x}^\star\in\mathbb{R}^n$ with minimum (in modulus) non-zero entry on the order of $\| \mathbf{x}^\star \|_2/\sqrt{k}$ from $k^2$ Gaussian measurements, modulo logarithmic terms. This yields a simple algorithm which, unlike most existing approaches to sparse phase retrieval, adapts to the sparsity level, without including thresholding steps or adding regularization terms. Our results also provide a principled theoretical understanding for Hadamard Wirtinger flow [54], as Euclidean gradient descent applied to the empirical risk problem with Hadamard parametrization can be recovered as a first-order approximation to mirror descent in discrete time.

PDF Details

ICML Conference 2020 Conference Paper

Decentralised Learning with Random Features and Distributed Gradient Descent

  • Dominic Richards
  • Patrick Rebeschini
  • Lorenzo Rosasco

We investigate the generalisation performance of Distributed Gradient Descent with implicit regularisation and random features in the homogenous setting where a network of agents are given data sampled independently from the same unknown distribution. Along with reducing the memory footprint, random features are particularly convenient in this setting as they provide a common parameterisation across agents that allows to overcome previous difficulties in implementing decentralised kernel regression. Under standard source and capacity assumptions, we establish high probability bounds on the predictive performance for each agent as a function of the step size, number of iterations, inverse spectral gap of the communication matrix and number of random features. By tuning these parameters, we obtain statistical rates that are minimax optimal with respect to the total number of samples in the network. The algorithm provides a linear improvement over single-machine gradient descent in memory cost and, when agents hold enough data with respect to the network size and inverse spectral gap, a linear speed up in computational run-time for any network topology. We present simulations that show how the number of random features, iterations and samples impact predictive performance.

Details

JMLR Journal 2020 Journal Article

Graph-Dependent Implicit Regularisation for Distributed Stochastic Subgradient Descent

  • Dominic Richards
  • Patrick Rebeschini

We propose graph-dependent implicit regularisation strategies for synchronised distributed stochastic subgradient descent (Distributed SGD) for convex problems in multi-agent learning. Under the standard assumptions of convexity, Lipschitz continuity, and smoothness, we establish statistical learning rates that retain, up to logarithmic terms, single-machine serial statistical guarantees through implicit regularisation (step size tuning and early stopping) with appropriate dependence on the graph topology. Our approach avoids the need for explicit regularisation in decentralised learning problems, such as adding constraints to the empirical risk minimisation rule. Particularly for distributed methods, the use of implicit regularisation allows the algorithm to remain simple, without projections or dual methods. To prove our results, we establish graph-independent generalisation bounds for Distributed SGD that match the single-machine serial SGD setting (using algorithmic stability), and we establish graph-dependent optimisation bounds that are of independent interest. We present numerical experiments to show that the qualitative nature of the upper bounds we derive can be representative of real behaviours. [abs] [ pdf ][ bib ] &copy JMLR 2020. ( edit, beta )

PDF Details

NeurIPS Conference 2020 Conference Paper

The Statistical Complexity of Early-Stopped Mirror Descent

  • Tomas Vaskevicius
  • Varun Kanade
  • Patrick Rebeschini

Recently there has been a surge of interest in understanding implicit regularization properties of iterative gradient-based optimization algorithms. In this paper, we study the statistical guarantees on the excess risk achieved by early-stopped unconstrained mirror descent algorithms applied to the unregularized empirical risk with the squared loss for linear models and kernel methods. By completing an inequality that characterizes convexity for the squared loss, we identify an intrinsic link between offset Rademacher complexities and potential-based convergence analysis of mirror descent methods. Our observation immediately yields excess risk guarantees for the path traced by the iterates of mirror descent in terms of offset complexities of certain function classes depending only on the choice of the mirror map, initialization point, step-size, and the number of iterations. We apply our theory to recover, in a rather clean and elegant manner via rather short proofs, some of the recent results in the implicit regularization literature, while also showing how to improve upon them in some settings.

PDF Details

JMLR Journal 2019 Journal Article

A New Approach to Laplacian Solvers and Flow Problems

  • Patrick Rebeschini
  • Sekhar Tatikonda

This paper investigates the behavior of the Min-Sum message passing scheme to solve systems of linear equations in the Laplacian matrices of graphs and to compute electric flows. Voltage and flow problems involve the minimization of quadratic functions and are fundamental primitives that arise in several domains. Algorithms that have been proposed are typically centralized and involve multiple graph-theoretic constructions or sampling mechanisms that make them difficult to implement and analyze. On the other hand, message passing routines are distributed, simple, and easy to implement. In this paper we establish a framework to analyze Min-Sum to solve voltage and flow problems. We characterize the error committed by the algorithm on general weighted graphs in terms of hitting times of random walks defined on the computation trees that support the operations of the algorithms with time. For $d$-regular graphs with equal weights, we show that the convergence of the algorithms is controlled by the total variation distance between the distributions of non-backtracking random walks defined on the original graph that start from neighboring nodes. The framework that we introduce extends the analysis of Min-Sum to settings where the contraction arguments previously considered in the literature (based on the assumption of walk summability or scaled diagonal dominance) can not be used, possibly in the presence of constraints. [abs] [ pdf ][ bib ] &copy JMLR 2019. ( edit, beta )

PDF Details

NeurIPS Conference 2019 Conference Paper

Decentralized Cooperative Stochastic Bandits

  • David Martínez-Rubio
  • Varun Kanade
  • Patrick Rebeschini

We study a decentralized cooperative stochastic multi-armed bandit problem with K arms on a network of N agents. In our model, the reward distribution of each arm is the same for each agent and rewards are drawn independently across agents and time steps. In each round, each agent chooses an arm to play and subsequently sends a message to her neighbors. The goal is to minimize the overall regret of the entire network. We design a fully decentralized algorithm that uses an accelerated consensus procedure to compute (delayed) estimates of the average of rewards obtained by all the agents for each arm, and then uses an upper confidence bound (UCB) algorithm that accounts for the delay and error of the estimates. We analyze the regret of our algorithm and also provide a lower bound. The regret is bounded by the optimal centralized regret plus a natural and simple term depending on the spectral gap of the communication matrix. Our algorithm is simpler to analyze than those proposed in prior work and it achieves better regret bounds, while requiring less information about the underlying network. It also performs better empirically.

PDF Details

NeurIPS Conference 2019 Conference Paper

Implicit Regularization for Optimal Sparse Recovery

  • Tomas Vaskevicius
  • Varun Kanade
  • Patrick Rebeschini

We investigate implicit regularization schemes for gradient descent methods applied to unpenalized least squares regression to solve the problem of reconstructing a sparse signal from an underdetermined system of linear measurements under the restricted isometry assumption. For a given parametrization yielding a non-convex optimization problem, we show that prescribed choices of initialization, step size and stopping time yield a statistically and computationally optimal algorithm that achieves the minimax rate with the same cost required to read the data up to poly-logarithmic factors. Beyond minimax optimality, we show that our algorithm adapts to instance difficulty and yields a dimension-independent rate when the signal-to-noise ratio is high enough. Key to the computational efficiency of our method is an increasing step size scheme that adapts to refined estimates of the true solution. We validate our findings with numerical experiments and compare our algorithm against explicit $\ell_{1}$ penalization. Going from hard instances to easy ones, our algorithm is seen to undergo a phase transition, eventually matching least squares with an oracle knowledge of the true support.

PDF Details

NeurIPS Conference 2019 Conference Paper

Optimal Statistical Rates for Decentralised Non-Parametric Regression with Linear Speed-Up

  • Dominic Richards
  • Patrick Rebeschini

We analyse the learning performance of Distributed Gradient Descent in the context of multi-agent decentralised non-parametric regression with the square loss function when i. i. d. samples are assigned to agents. We show that if agents hold sufficiently many samples with respect to the network size, then Distributed Gradient Descent achieves optimal statistical rates with a number of iterations that scales, up to a threshold, with the inverse of the spectral gap of the gossip matrix divided by the number of samples owned by each agent raised to a problem-dependent power. The presence of the threshold comes from statistics. It encodes the existence of a "big data" regime where the number of required iterations does not depend on the network topology. In this regime, Distributed Gradient Descent achieves optimal statistical rates with the same order of iterations as gradient descent run with all the samples in the network. Provided the communication delay is sufficiently small, the distributed protocol yields a linear speed-up in runtime compared to the single-machine protocol. This is in contrast to decentralised optimisation algorithms that do not exploit statistics and only yield a linear speed-up in graphs where the spectral gap is bounded away from zero. Our results exploit the statistical concentration of quantities held by agents and shed new light on the interplay between statistics and communication in decentralised methods. Bounds are given in the standard non-parametric setting with source/capacity assumptions.

PDF Details

NeurIPS Conference 2017 Conference Paper

Accelerated consensus via Min-Sum Splitting

  • Patrick Rebeschini
  • Sekhar Tatikonda

We apply the Min-Sum message-passing protocol to solve the consensus problem in distributed optimization. We show that while the ordinary Min-Sum algorithm does not converge, a modified version of it known as Splitting yields convergence to the problem solution. We prove that a proper choice of the tuning parameters allows Min-Sum Splitting to yield subdiffusive accelerated convergence rates, matching the rates obtained by shift-register methods. The acceleration scheme embodied by Min-Sum Splitting for the consensus problem bears similarities with lifted Markov chains techniques and with multi-step first order methods in convex optimization.

PDF Details