Here we revisit the classic problem of linear quadratic estimation, i.e. estimating the trajectory of a linear dynamical system from noisy measurements. The celebrated Kalman filter gives an optimal estimator when the measurement noise is Gaussian, but is widely known to break down when one deviates from this assumption, e.g. when the noise is heavy-tailed. Many ad hoc heuristics have been employed in practice for dealing with outliers. In a pioneering work, Schick and Mitter gave provable guarantees when the measurement noise is a known infinitesimal perturbation of a Gaussian and raised the important question of whether one can get similar guarantees for large and unknown perturbations. In this work we give a truly robust filter: we give the first strong provable guarantees for linear quadratic estimation when even a constant fraction of measurements have been adversarially corrupted. This framework can model heavy-tailed and even non-stationary noise processes. Our algorithm robustifies the Kalman filter in the sense that it competes with the optimal algorithm that knows the locations of the corruptions. Our work is in a challenging Bayesian setting where the number of measurements scales with the complexity of what we need to estimate. Moreover, in linear dynamical systems past information decays over time. We develop a suite of new techniques to robustly extract information across different time steps and over varying time scales.

Despite rapid progress in theoretical reinforcement learning (RL) over the last few years, most of the known guarantees are worst-case in nature, failing to take advantage of structure that may be known a priori about a given RL problem at hand. In this paper we address the question of whether worst-case lower bounds for regret in online learning of Markov decision processes (MDPs) can be circumvented when information about the MDP, in the form of predictions about its optimal $Q$-value function, is given to the algorithm. We show that when the predictions about the optimal $Q$-value function satisfy a reasonably weak condition we call distillation, then we can improve regret bounds by replacing the set of state-action pairs with the set of state-action pairs on which the predictions are grossly inaccurate. This improvement holds for both uniform regret bounds and gap-based ones. Further, we are able to achieve this property with an algorithm that achieves sublinear regret when given arbitrary predictions (i.e., even those which are not a distillation). Our work extends a recent line of work on algorithms with predictions, which has typically focused on simple online problems such as caching and scheduling, to the more complex and general problem of reinforcement learning.

Many works in signal processing and learning theory operate under the assumption that the underlying model is simple, e.g. that a signal is approximately $k$-Fourier-sparse or that a distribution can be approximated by a mixture model that has at most $k$ components. However the problem of fitting the parameters of such a model becomes more challenging when the frequencies/components are too close together. In this work we introduce new methods for sparsifying sums of exponentials and give various algorithmic applications. First we study Fourier-sparse interpolation without a frequency gap, where Chen et al. gave an algorithm for finding an $\epsilon$-approximate solution which uses $k' = \mbox{poly}(k, \log 1/\epsilon)$ frequencies. Second, we study learning Gaussian mixture models in one dimension without a separation condition. Kernel density estimators give an $\epsilon$-approximation that uses $k' = O(k/\epsilon^2)$ components. These methods both output models that are much more complex than what we started out with. We show how to post-process to reduce the number of frequencies/components down to $k' = \widetilde{O}(k)$, which is optimal up to logarithmic factors. Moreover we give applications to model selection. In particular, we give the first algorithms for approximately (and robustly) determining the number of components in a Gaussian mixture model that work without a separation condition.

In this work we study the orbit recovery problem, which is a natural abstraction for the problem of recovering a planted signal from noisy measurements under unknown group actions. Many important inverse problems in statistics, engineering and the sciences fit into this framework. Prior work has studied cases when the group is discrete and/or abelian. However fundamentally new techniques are needed in order to handle more complex group actions. Our main result is a quasi-polynomial time algorithm to solve orbit recovery over $SO(3)$ - i.e. the cryo-electron tomography problem which asks to recover the three-dimensional structure of a molecule from noisy measurements of randomly rotated copies of it. We analyze a variant of the frequency marching heuristic in the framework of smoothed analysis. Our approach exploits the layered structure of the invariant polynomials, and simultaneously yields a new class of tensor decomposition algorithms that work in settings when the tensor is not low-rank but rather where the factors are algebraically related to each other by a group action.

Gaussian mixture models have a long and storied history. They were first introduced in a groundbreaking work of Karl Pearson[33] in 1894 and have found wide-ranging applications ever since, as a natural model for data believed to be coming from two or more heterogeneous sources. Early works focused on the statistical complexity [35], namely bounding the number of samples needed to estimate the Gaussian mixture model to within some desired accuracy. More recently, these problems have been revisited with the emphasis being on giving computationally efficient algorithms that work in high dimensions and with minimal assumptions [8,11,21,25,32]. There are different types of learning goals we could ask for, and the distinctions between them will play an important role in understanding the context of our work: (1) In parameter learning, we want to estimate the mixture on a component-by-component basis. We ask that there is a matching between the components in our hypothesis and those of the true mixture so that across the matching we are close in total variation distance. Alternatively we could ask to be close in an appropriate parameter distance instead.

Geodesically convex optimization is a natural generalization that replaces Euclidean space with a Riemannian manifold and we require that the function we want to minimize is convex along geodesics [1, 5, 31]. It turns out that many optimization problems of interest, while non-convex in the Euclidean view, become geodesically convex when equipped with the right geometry. Some notable examples: The fastest known algorithms for computing Brascamp-Lieb constants [3, 17], and solving related problems like the null cone problem [6-8], exploit geodesic convexity. In machine learning, it arises in matrix completion [9, 28, 33], dictionary learning [11, 26], robust subspace recovery [39], mixture models [18] and optimization under orthogonality constraints [14]. In statistics, some basic problems like estimating the shape of an elliptical distribution [16, 35] or estimation matrix normal models [4, 29] are best viewed through the lens of geodesic convexity.

This work represents a natural coalescence of two important lines of work: learning mixtures of Gaussians and algorithmic robust statistics. In particular we give the first provably robust algorithm for learning mixtures of any constant number of Gaussians. We require only mild assumptions on the mixing weights (bounded fractionality) and that the total variation distance between components is bounded away from zero. At the heart of our algorithm is a new method for proving dimension-independent polynomial identifiability through applying a carefully chosen sequence of differential operations to certain generating functions that not only encode the parameters we would like to learn but also the system of polynomial equations we would like to solve. We show how the symbolic identities we derive can be directly used to analyze a natural sum-of-squares relaxation.

In this work we revisit two classic high-dimensional online learning problems, namely regression and linear contextual bandits, from the perspective of adversarial robustness. Existing works in algorithmic robust statistics make strong distributional assumptions that ensure that the input data is evenly spread out or comes from a nice generative model. Is it possible to achieve strong robustness guarantees even without distributional assumptions altogether, where the sequence of tasks we are asked to solve is adaptively and adversarially chosen? We answer this question in the affirmative for both regression and linear contextual bandits. In fact our algorithms succeed where convex surrogates fail in the sense that we show strong lower bounds categorically for the existing approaches. Our approach is based on a novel way to use the sum-of-squares hierarchy in online learning and in the absence of distributional assumptions. Moreover we give extensions of our main results to infinite dimensional settings where the feature vectors are represented implicitly via a kernel map.

In this paper we revisit some classic problems on classification under misspecification. In particular, we study the problem of learning halfspaces under Massart noise with rate $\eta$. In a recent work, Diakonikolas, Goulekakis, and Tzamos resolved a long-standing problem by giving the first efficient algorithm for learning to accuracy $\eta + \epsilon$ for any $\epsilon > 0$. However, their algorithm outputs a complicated hypothesis, which partitions space into $\text{poly}(d,1/\epsilon)$ regions. Here we give a much simpler algorithm and in the process resolve a number of outstanding open questions: (1) We give the first proper learner for Massart halfspaces that achieves $\eta + \epsilon$. We also give improved bounds on the sample complexity achievable by polynomial time algorithms. (2) Based on (1), we develop a blackbox knowledge distillation procedure to convert an arbitrarily complex classifier to an equally good proper classifier. (3) By leveraging a simple but overlooked connection to evolvability, we show any SQ algorithm requires super-polynomially many queries to achieve $\mathsf{OPT} + \epsilon$. Moreover we study generalized linear models where $\mathbb{E}[Y|\mathbf{X}] = \sigma(\langle \mathbf{w}^*, \mathbf{X}\rangle)$ for any odd, monotone, and Lipschitz function $\sigma$. This family includes the previously mentioned halfspace models as a special case, but is much richer and includes other fundamental models like logistic regression. We introduce a challenging new corruption model that generalizes Massart noise, and give a general algorithm for learning in this setting. Our algorithms are based on a small set of core recipes for learning to classify in the presence of misspecification. Finally we study our algorithm for learning halfspaces under Massart noise empirically and find that it exhibits some appealing fairness properties.

Gaussian Graphical Models (GGMs) have wide-ranging applications in machine learning and the natural and social sciences. In most of the settings in which they are applied, the number of observed samples is much smaller than the dimension and they are assumed to be sparse. While there are a variety of algorithms (e.g. Graphical Lasso, CLIME) that provably recover the graph structure with a logarithmic number of samples, they assume various conditions that require the precision matrix to be in some sense well-conditioned. Here we give the first polynomial-time algorithms for learning attractive GGMs and walk-summable GGMs with a logarithmic number of samples without any such assumptions. In particular, our algorithms can tolerate strong dependencies among the variables. We complement our results with experiments showing that many existing algorithms fail even in some simple settings where there are long dependency chains, whereas ours do not.