We provide a new adaptive method for online convex optimization, MetaGrad, that is robust to general convex losses but achieves faster rates for a broad class of special functions, including exp-concave and strongly convex functions, but also various types of stochastic and non-stochastic functions without any curvature. We prove this by drawing a connection to the Bernstein condition, which is known to imply fast rates in offline statistical learning. MetaGrad further adapts automatically to the size of the gradients. Its main feature is that it simultaneously considers multiple learning rates, which are weighted directly proportional to their empirical performance on the data using a new meta-algorithm. We provide three versions of MetaGrad. The full matrix version maintains a full covariance matrix and is applicable to learning tasks for which we can afford update time quadratic in the dimension. The other two versions provide speed-ups for high-dimensional learning tasks with an update time that is linear in the dimension: one is based on sketching, the other on running a separate copy of the basic algorithm per coordinate. We evaluate all versions of MetaGrad on benchmark online classification and regression tasks, on which they consistently outperform both online gradient descent and AdaGrad.

We propose a novel approach to analyze generalization error for discretizations of Langevin diffusion, such as the stochastic gradient Langevin dynamics (SGLD). For an $\epsilon$ tolerance of expected generalization error, it is known that a first order discretization can reach this target if we run $\Omega(\epsilon^{-1} \log (\epsilon^{-1}) )$ iterations with $\Omega(\epsilon^{-1})$ samples. In this article, we show that with additional smoothness assumptions, even first order methods can achieve arbitrarily runtime complexity. More precisely, for each $N>0$, we provide a sufficient smoothness condition on the loss function such that a first order discretization can reach $\epsilon$ expected generalization error given $\Omega( \epsilon^{-1/N} \log (\epsilon^{-1}) )$ iterations with $\Omega(\epsilon^{-1})$ samples.

We model the dynamics of privacy loss in Langevin diffusion and extend it to the noisy gradient descent algorithm: we compute a tight bound on R\'enyi differential privacy and the rate of its change throughout the learning process. We prove that the privacy loss converges exponentially fast. This significantly improves the prior privacy analysis of differentially private (stochastic) gradient descent algorithms, where (R\'enyi) privacy loss constantly increases over the training iterations. Unlike composition-based methods in differential privacy, our privacy analysis does not assume that the noisy gradients (or parameters) during the training could be revealed to the adversary. Our analysis tracks the dynamics of privacy loss through the algorithm's intermediate parameter distributions, thus allowing us to account for privacy amplification due to convergence. We prove that our privacy analysis is tight, and also provide a utility analysis for strongly convex, smooth and Lipshitz loss functions.

As machine learning models are deployed in critical applications, it becomes important to not just provide point estimators of the model parameters (or subsequent predictions), but also quantify the uncertainty associated with estimating the model parameters via confidence sets. In the last decade, estimating or training in several machine learning models has become synonymous with running stochastic gradient algorithms. However, computing the stochastic gradients in several settings is highly expensive or even impossible at times. An important question which has thus far not been addressed sufficiently in the statistical machine learning literature is that of equipping zeroth-order stochastic gradient algorithms with practical yet rigorous inferential capabilities. Towards this, in this work, we first establish a central limit theorem for Polyak-Ruppert averaged stochastic gradient algorithm in the zeroth-order setting. We then provide online estimators of the asymptotic covariance matrix appearing in the central limit theorem, thereby providing a practical procedure for constructing asymptotically valid confidence sets (or intervals) for parameter estimation (or prediction) in the zeroth-order setting.

The noise in stochastic gradient descent (SGD), caused by minibatch sampling, remains poorly understood despite its enormous practical importance in offering good training efficiency and generalization ability. In this work, we study the minibatch noise in SGD. Motivated by the observation that minibatch sampling does not always cause a fluctuation, we set out to find the conditions that cause minibatch noise to emerge. We first derive the analytically solvable results for linear regression under various settings, which are compared to the commonly used approximations that are used to understand SGD noise. We show that some degree of mismatch between model and data complexity is needed in order for SGD to "cause" a noise, and that such mismatch may be due to the existence of static noise in the labels, in the input, the use of regularization, or underparametrization. Our results motivate a more accurate general formulation to describe minibatch noise.

Stochastic Gradient Descent (SGD) based methods have been widely used for training large-scale machine learning models that also generalize well in practice. Several explanations have been offered for this generalization performance, a prominent one being algorithmic stability [18]. However, there are no known examples of smooth loss functions for which the analysis can be shown to be tight. Furthermore, apart from the properties of the loss function, data distribution has also been shown to be an important factor in generalization performance. This raises the question: is the stability analysis of [18] tight for smooth functions, and if not, for what kind of loss functions and data distributions can the stability analysis be improved? In this paper we first settle open questions regarding tightness of bounds in the data-independent setting: we show that for general datasets, the existing analysis for convex and strongly-convex loss functions is tight, but it can be improved for non-convex loss functions. Next, we give a novel and improved data-dependent bounds: we show stability upper bounds for a large class of convex regularized loss functions, with negligible regularization parameters, and improve existing data-dependent bounds in the non-convex setting. We hope that our results will initiate further efforts to better understand the data-dependent setting under non-convex loss functions, leading to an improved understanding of the generalization abilities of deep networks.

The automation of probabilistic reasoning is one of the primary aims of machine learning. Recently, the confluence of variational inference and deep learning has led to powerful and flexible automatic inference methods that can be trained by stochastic gradient descent. In particular, normalizing flows are highly parameterized deep models that can fit arbitrarily complex posterior densities. However, normalizing flows struggle in highly structured probabilistic programs as they need to relearn the forward-pass of the program. Automatic structured variational inference (ASVI) remedies this problem by constructing variational programs that embed the forward-pass. Here, we combine the flexibility of normalizing flows and the prior-embedding property of ASVI in a new family of variational programs, which we named cascading flows. A cascading flows program interposes a newly designed highway flow architecture in between the conditional distributions of the prior program such as to steer it toward the observed data. These programs can be constructed automatically from an input probabilistic program and can also be amortized automatically. We evaluate the performance of the new variational programs in a series of structured inference problems. We find that cascading flows have much higher performance than both normalizing flows and ASVI in a large set of structured inference problems.

Stochastic bilevel optimization generalizes the classic stochastic optimization from the minimization of a single objective to the minimization of an objective function that depends the solution of another optimization problem. Recently, stochastic bilevel optimization is regaining popularity in emerging machine learning applications such as hyper-parameter optimization and model-agnostic meta learning. To solve this class of stochastic optimization problems, existing methods require either double-loop or two-timescale updates, which are sometimes less efficient. This paper develops a new optimization method for a class of stochastic bilevel problems that we term Single-Timescale stochAstic BiLevEl optimization (STABLE) method. STABLE runs in a single loop fashion, and uses a single-timescale update with a fixed batch size. To achieve an $\epsilon$-stationary point of the bilevel problem, STABLE requires ${\cal O}(\epsilon^{-2})$ samples in total; and to achieve an $\epsilon$-optimal solution in the strongly convex case, STABLE requires ${\cal O}(\epsilon^{-1})$ samples. To the best of our knowledge, this is the first bilevel optimization algorithm achieving the same order of sample complexity as the stochastic gradient descent method for the single-level stochastic optimization.

We study the convergence rate of continuous-time simulated annealing $(X_t; \, t \ge 0)$ and its discretization $(x_k; \, k =0,1, \ldots)$ for approximating the global optimum of a given function $f$. We prove that the tail probability $\mathbb{P}(f(X_t) > \min f +\delta)$ (resp. $\mathbb{P}(f(x_k) > \min f +\delta)$) decays polynomial in time (resp. in cumulative step size), and provide an explicit rate as a function of the model parameters. Our argument applies the recent development on functional inequalities for the Gibbs measure at low temperatures -- the Eyring-Kramers law. In the discrete setting, we obtain a condition on the step size to ensure the convergence.

The stability and generalization of stochastic gradient-based methods provide valuable insights into understanding the algorithmic performance of machine learning models. As the main workhorse for deep learning, stochastic gradient descent has received a considerable amount of studies. Nevertheless, the community paid little attention to its decentralized variants. In this paper, we provide a novel formulation of the decentralized stochastic gradient descent. Leveraging this formulation together with (non)convex optimization theory, we establish the first stability and generalization guarantees for the decentralized stochastic gradient descent. Our theoretical results are built on top of a few common and mild assumptions and reveal that the decentralization deteriorates the stability of SGD for the first time. We verify our theoretical findings by using a variety of decentralized settings and benchmark machine learning models.