arXiv subjects

Aaron Sidford

Publications and source records attributed to Aaron Sidford.

At least 91 records · Page 5Linked to original sources

arXiv2020

Memory-Sample Tradeoffs for Linear Regression with Small Error

We consider the problem of performing linear regression over a stream of $d$-dimensional examples, and show that any algorithm that uses a subquadratic amount of memory exhibits a slower rate of convergence than can be achieved without memory constraints. Specifically, consider a sequence of labeled examples $(a_1,b_1), (a_2,b_2)\ldots,$ with $a_i$ drawn independently from a $d$-dimensional isotropic Gaussian, and where $b_i = \langle a_i, x\rangle + η_i,$ for a fixed $x \in \mathbb{R}^d$ with $\|x\|_2 = 1$ and with independent noise $η_i$ drawn uniformly from the interval $[-2^{-d/5},2^{-d/5}].$ We show that any algorithm with at most $d^2/4$ bits of memory requires at least $Ω(d \log \log \frac{1}ε)$ samples to approximate $x$ to $\ell_2$ error $ε$ with probability of success at least $2/3$, for $ε$ sufficiently small as a function of $d$. In contrast, for such $ε$, $x$ can be recovered to error $ε$ with probability $1-o(1)$ with memory $O\left(d^2 \log(1/ε)\right)$ using $d$ examples. This represents the first nontrivial lower bounds for regression with super-linear memory, and may open the door for strong memory/sample tradeoffs for continuous optimization.

cs.LG↗

arXiv2020

Coordinate Methods for Matrix Games

We develop primal-dual coordinate methods for solving bilinear saddle-point problems of the form $\min_{x \in \mathcal{X}} \max_{y\in\mathcal{Y}} y^\top A x$ which contain linear programming, classification, and regression as special cases. Our methods push existing fully stochastic sublinear methods and variance-reduced methods towards their limits in terms of per-iteration complexity and sample complexity. We obtain nearly-constant per-iteration complexity by designing efficient data structures leveraging Taylor approximations to the exponential and a binomial heap. We improve sample complexity via low-variance gradient estimators using dynamic sampling distributions that depend on both the iterates and the magnitude of the matrix entries. Our runtime bounds improve upon those of existing primal-dual methods by a factor depending on sparsity measures of the $m$ by $n$ matrix $A$. For example, when rows and columns have constant $\ell_1/\ell_2$ norm ratios, we offer improvements by a factor of $m+n$ in the fully stochastic setting and $\sqrt{m+n}$ in the variance-reduced setting. We apply our methods to computational geometry problems, i.e. minimum enclosing ball, maximum inscribed ball, and linear regression, and obtain improved complexity bounds. For linear regression with an elementwise nonnegative matrix, our guarantees improve on exact gradient methods by a factor of $\sqrt{\mathrm{nnz}(A)/(m+n)}$.

cs.DS↗

arXiv2020

Solving Linear Programs with Sqrt(rank) Linear System Solves

We present an algorithm that given a linear program with $n$ variables, $m$ constraints, and constraint matrix $A$, computes an $ε$-approximate solution in $\tilde{O}(\sqrt{rank(A)}\log(1/ε))$ iterations with high probability. Each iteration of our method consists of solving $\tilde{O}(1)$ linear systems and additional nearly linear time computation, improving by a factor of $\tildeΩ((m/rank(A))^{1/2})$ over the previous fastest method with this iteration cost due to Renegar (1988). Further, we provide a deterministic polynomial time computable $\tilde{O}(rank(A))$-self-concordant barrier function for the polytope, resolving an open question of Nesterov and Nemirovski (1994) on the theory of "universal barriers" for interior point methods. Applying our techniques to the linear program formulation of maximum flow yields an $\tilde{O}(|E|\sqrt{|V|}\log(U))$ time algorithm for solving the maximum flow problem on directed graphs with $|E|$ edges, $|V|$ vertices, and integer capacities of size at most $U$. This improves upon the previous fastest polynomial running time of $O(|E|\min\{|E|^{1/2},|V|^{2/3}\}\log(|V|^{2}/|E|)\log(U))$ achieved by Goldberg and Rao (1998). In the special case of solving dense directed unit capacity graphs our algorithm improves upon the previous fastest running times of $O(|E|\min\{|E|^{1/2},|V|^{2/3}\})$ achieved by Even and Tarjan (1975) and Karzanov (1973) and of $\tilde{O}(|E|^{10/7})$ achieved more recently by Mądry (2013).

cs.DS↗

arXiv2020

Efficiently Solving MDPs with Stochastic Mirror Descent

We present a unified framework based on primal-dual stochastic mirror descent for approximately solving infinite-horizon Markov decision processes (MDPs) given a generative model. When applied to an average-reward MDP with $A_{tot}$ total state-action pairs and mixing time bound $t_{mix}$ our method computes an $ε$-optimal policy with an expected $\widetilde{O}(t_{mix}^2 A_{tot} ε^{-2})$ samples from the state-transition matrix, removing the ergodicity dependence of prior art. When applied to a $γ$-discounted MDP with $A_{tot}$ total state-action pairs our method computes an $ε$-optimal policy with an expected $\widetilde{O}((1-γ)^{-4} A_{tot} ε^{-2})$ samples, matching the previous state-of-the-art up to a $(1-γ)^{-1}$ factor. Both methods are model-free, update state values and policies simultaneously, and run in time linear in the number of samples taken. We achieve these results through a more general stochastic mirror descent framework for solving bilinear saddle-point problems with simplex and box domains and we demonstrate the flexibility of this framework by providing further applications to constrained MDPs.

cs.LG↗

arXiv2020

Constant Girth Approximation for Directed Graphs in Subquadratic Time

In this paper we provide a $\tilde{O}(m\sqrt{n})$ time algorithm that computes a $3$-multiplicative approximation of the girth of a $n$-node $m$-edge directed graph with non-negative edge lengths. This is the first algorithm which approximates the girth of a directed graph up to a constant multiplicative factor faster than All-Pairs Shortest Paths (APSP) time, i.e. $O(mn)$. Additionally, for any integer $k \ge 1$, we provide a deterministic algorithm for a $O(k\log\log n)$-multiplicative approximation to the girth in directed graphs in $\tilde{O}(m^{1+1/k})$ time. Combining the techniques from these two results gives us an algorithm for a $O(k\log k)$-multiplicative approximation to the girth in directed graphs in $\tilde{O}(m^{1+1/k})$ time. Our results naturally also provide algorithms for improved constructions of roundtrip spanners, the analog of spanners in directed graphs. The previous fastest algorithms for these problems either ran in All-Pairs Shortest Paths (APSP) time, i.e. $O(mn)$, or were due Pachocki et al. (PRSTV18) which provided a randomized algorithm that for any integer $k \ge 1$ in time $\tilde{O}(m^{1+1/k})$ computed with high probability a $O(k\log n)$ multiplicative approximation of the girth. Our first algorithm constitutes the first sub-APSP-time algorithm for approximating the girth to constant accuracy, our second removes the need for randomness and improves the approximation factor in Pachocki et al. (PRSTV18), and our third is the first time versus quality trade-off for obtaining constant approximations.

cs.DS↗

arXiv2020

Faster Divergence Maximization for Faster Maximum Flow

In this paper we provide an algorithm which given any $m$-edge $n$-vertex directed graph with integer capacities at most $U$ computes a maximum $s$-$t$ flow for any vertices $s$ and $t$ in $m^{4/3+o(1)}U^{1/3}$ time. This improves upon the previous best running times of $m^{11/8+o(1)}U^{1/4}$ (Liu Sidford 2019), $\tilde{O}(m \sqrt{n} \log U)$ (Lee Sidford 2014), and $O(mn)$ (Orlin 2013) when the graph is not too dense or has large capacities. To achieve the results this paper we build upon previous algorithmic approaches to maximum flow based on interior point methods (IPMs). In particular, we overcome a key bottleneck of previous advances in IPMs for maxflow (Mądry 2013, Mądry 2016, Liu Sidford 2019), which make progress by maximizing the energy of local $\ell_2$ norm minimizing electric flows. We generalize this approach and instead maximize the divergence of flows which minimize the Bregman divergence distance with respect to the weighted logarithmic barrier. This allows our algorithm to avoid dependencies on the $\ell_4$ norm that appear in other IPM frameworks (e.g. Cohen Mądry Sankowski Vladu 2017, Axiotis Mądry Vladu 2020). Further, we show that smoothed $\ell_2$-$\ell_p$ flows (Kyng, Peng, Sachdeva, Wang 2019), which we previously used to efficiently maximize energy (Liu Sidford 2019), can also be used to efficiently maximize divergence, thereby yielding our desired runtimes. We believe both this generalized view of energy maximization and generalized flow solvers we develop may be of further interest.

cs.DS↗

arXiv2020

The Bethe and Sinkhorn Permanents of Low Rank Matrices and Implications for Profile Maximum Likelihood

In this paper we consider the problem of computing the likelihood of the profile of a discrete distribution, i.e., the probability of observing the multiset of element frequencies, and computing a profile maximum likelihood (PML) distribution, i.e., a distribution with the maximum profile likelihood. For each problem we provide polynomial time algorithms that given $n$ i.i.d.\ samples from a discrete distribution, achieve an approximation factor of $\exp\left(-O(\sqrt{n} \log n) \right)$, improving upon the previous best-known bound achievable in polynomial time of $\exp(-O(n^{2/3} \log n))$ (Charikar, Shiragur and Sidford, 2019). Through the work of Acharya, Das, Orlitsky and Suresh (2016), this implies a polynomial time universal estimator for symmetric properties of discrete distributions in a broader range of error parameter. We achieve these results by providing new bounds on the quality of approximation of the Bethe and Sinkhorn permanents (Vontobel, 2012 and 2014). We show that each of these are $\exp(O(k \log(N/k)))$ approximations to the permanent of $N \times N$ matrices with non-negative rank at most $k$, improving upon the previous known bounds of $\exp(O(N))$. To obtain our results on PML, we exploit the fact that the PML objective is proportional to the permanent of a certain Vandermonde matrix with $\sqrt{n}$ distinct columns, i.e. with non-negative rank at most $\sqrt{n}$. As a by-product of our work we establish a surprising connection between the convex relaxation in prior work (CSS19) and the well-studied Bethe and Sinkhorn approximations.

cs.DS↗

arXiv2020

Coordinate Methods for Accelerating $\ell_\infty$ Regression and Faster Approximate Maximum Flow

We provide faster algorithms for approximately solving $\ell_{\infty}$ regression, a fundamental problem prevalent in both combinatorial and continuous optimization. In particular, we provide accelerated coordinate descent methods capable of provably exploiting dynamic measures of coordinate smoothness, and apply them to $\ell_\infty$ regression over a box to give algorithms which converge in $k$ iterations at a $O(1/k)$ rate. Our algorithms can be viewed as an alternative approach to the recent breakthrough result of Sherman [She17] which achieves a similar runtime improvement over classic algorithmic approaches, i.e. smoothing and gradient descent, which either converge at a $O(1/\sqrt{k})$ rate or have running times with a worse dependence on problem parameters. Our runtimes match those of [She17] across a broad range of parameters and achieve improvement in certain structured cases. We demonstrate the efficacy of our result by providing faster algorithms for the well-studied maximum flow problem. Directly leveraging our accelerated $\ell_\infty$ regression algorithms imply a $\tilde{O}\left(m + \sqrt{mn}/ε\right)$ runtime to compute an $ε$-approximate maximum flow for an undirected graph with $m$ edges and $n$ vertices, generically improving upon the previous best known runtime of $\tilde{O}\left(m/ε\right)$ in [She17] whenever the graph is slightly dense. We further design an algorithm adapted to the structure of the regression problem induced by maximum flow obtaining a runtime of $\tilde{O}\left(m + \max(n, \sqrt{ns})/ε\right)$, where $s$ is the squared $\ell_2$ norm of the congestion of any optimal flow. Moreover, we show how to leverage this result to achieve improved exact algorithms for maximum flow on a variety of unit capacity graphs. We hope that our work serves as an important step towards achieving even faster maximum flow algorithms.

cs.DS↗

arXiv2020

Acceleration with a Ball Optimization Oracle

Consider an oracle which takes a point $x$ and returns the minimizer of a convex function $f$ in an $\ell_2$ ball of radius $r$ around $x$. It is straightforward to show that roughly $r^{-1}\log\frac{1}ε$ calls to the oracle suffice to find an $ε$-approximate minimizer of $f$ in an $\ell_2$ unit ball. Perhaps surprisingly, this is not optimal: we design an accelerated algorithm which attains an $ε$-approximate minimizer with roughly $r^{-2/3} \log \frac{1}ε$ oracle queries, and give a matching lower bound. Further, we implement ball optimization oracles for functions with locally stable Hessians using a variant of Newton's method. The resulting algorithm applies to a number of problems of practical and theoretical import, improving upon previous results for logistic and $\ell_\infty$ regression and achieving guarantees comparable to the state-of-the-art for $\ell_p$ regression.

math.OC↗

arXiv2020

A General Framework for Symmetric Property Estimation

In this paper we provide a general framework for estimating symmetric properties of distributions from i.i.d. samples. For a broad class of symmetric properties we identify the easy region where empirical estimation works and the difficult region where more complex estimators are required. We show that by approximately computing the profile maximum likelihood (PML) distribution \cite{ADOS16} in this difficult region we obtain a symmetric property estimation framework that is sample complexity optimal for many properties in a broader parameter regime than previous universal estimation approaches based on PML. The resulting algorithms based on these pseudo PML distributions are also more practical.

cs.DS↗

arXiv2020

Towards Optimal Running Times for Optimal Transport

In this work, we provide faster algorithms for approximating the optimal transport distance, e.g. earth mover's distance, between two discrete probability distributions $μ, ν\in Δ^n$. Given a cost function $C : [n] \times [n] \to \mathbb{R}_{\geq 0}$ where $C(i,j) \leq 1$ quantifies the penalty of transporting a unit of mass from $i$ to $j$, we show how to compute a coupling $X$ between $r$ and $c$ in time $\widetilde{O}\left(n^2 /ε\right)$ whose expected transportation cost is within an additive $ε$ of optimal. This improves upon the previously best known running time for this problem of $\widetilde{O}\left(\text{min}\left\{ n^{9/4}/ε, n^2/ε^2 \right\}\right)$. We achieve our results by providing reductions from optimal transport to canonical optimization problems for which recent algorithmic efforts have provided nearly-linear time algorithms. Leveraging nearly linear time algorithms for solving packing linear programs and for solving the matrix balancing problem, we obtain two separate proofs of our stated running time. Further, one of our algorithms is easily parallelized and can be implemented with depth $\widetilde{O}(1/ε)$. Moreover, we show that further algorithmic improvements to our result would be surprising in the sense that any improvement would yield an $o(n^{2.5})$ algorithm for \textit{maximum cardinality bipartite matching}, for which currently the only known algorithms for achieving such a result are based on fast-matrix multiplication.

cs.DS↗

arXiv2019

Parallel Reachability in Almost Linear Work and Square Root Depth

In this paper we provide a parallel algorithm that given any $n$-node $m$-edge directed graph and source vertex $s$ computes all vertices reachable from $s$ with $\tilde{O}(m)$ work and $n^{1/2 + o(1)}$ depth with high probability in $n$ . This algorithm also computes a set of $\tilde{O}(n)$ edges which when added to the graph preserves reachability and ensures that the diameter of the resulting graph is at most $n^{1/2 + o(1)}$. Our result improves upon the previous best known almost linear work reachability algorithm due to Fineman which had depth $\tilde{O}(n^{2/3})$. Further, we show how to leverage this algorithm to achieve improved distributed algorithms for single source reachability in the CONGEST model. In particular, we provide a distributed algorithm that given a $n$-node digraph of undirected hop-diameter $D$ solves the single source reachability problem with $\tilde{O}(n^{1/2} + n^{1/3 + o(1)} D^{2/3})$ rounds of the communication in the CONGEST model with high probability in $n$. Our algorithm is nearly optimal whenever $D = O(n^{1/4 - ε})$ for any constant $ε> 0$ and is the first nearly optimal algorithm for general graphs whose diameter is $Ω(n^δ)$ for any constant $δ$.

cs.DS↗

arXiv2019

Variance Reduction for Matrix Games

We present a randomized primal-dual algorithm that solves the problem $\min_{x} \max_{y} y^\top A x$ to additive error $ε$ in time $\mathrm{nnz}(A) + \sqrt{\mathrm{nnz}(A)n}/ε$, for matrix $A$ with larger dimension $n$ and $\mathrm{nnz}(A)$ nonzero entries. This improves the best known exact gradient methods by a factor of $\sqrt{\mathrm{nnz}(A)/n}$ and is faster than fully stochastic gradient methods in the accurate and/or sparse regime $ε\le \sqrt{n/\mathrm{nnz}(A)}$. Our results hold for $x,y$ in the simplex (matrix games, linear programming) and for $x$ in an $\ell_2$ ball and $y$ in the simplex (perceptron / SVM, minimum enclosing ball). Our algorithm combines Nemirovski's "conceptual prox-method" and a novel reduced-variance gradient estimator based on "sampling from the difference" between the current iterate and a reference point.

math.OC↗

arXiv2019

Principal Component Projection Without Principal Component Analysis

We show how to efficiently project a vector onto the top principal components of a matrix, without explicitly computing these components. Specifically, we introduce an iterative algorithm that provably computes the projection using few calls to any black-box routine for ridge regression. By avoiding explicit principal component analysis (PCA), our algorithm is the first with no runtime dependence on the number of top principal components. We show that it can be used to give a fast iterative method for the popular principal component regression problem, giving the first major runtime improvement over the naive method of combining PCA with regression. To achieve our results, we first observe that ridge regression can be used to obtain a "smooth projection" onto the top principal components. We then sharpen this approximation to true projection using a low-degree polynomial approximation to the matrix step function. Step function approximation is a topic of long-term interest in scientific computing. We extend prior theory by constructing polynomials with simple iterative structure and rigorously analyzing their behavior under limited precision.

cs.DS↗

arXiv2019

Deterministic Approximation of Random Walks in Small Space

We give a deterministic, nearly logarithmic-space algorithm that given an undirected graph $G$, a positive integer $r$, and a set $S$ of vertices, approximates the conductance of $S$ in the $r$-step random walk on $G$ to within a factor of $1+ε$, where $ε>0$ is an arbitrarily small constant. More generally, our algorithm computes an $ε$-spectral approximation to the normalized Laplacian of the $r$-step walk. Our algorithm combines the derandomized square graph operation (Rozenman and Vadhan, 2005), which we recently used for solving Laplacian systems in nearly logarithmic space (Murtagh, Reingold, Sidford, and Vadhan, 2017), with ideas from (Cheng, Cheng, Liu, Peng, and Teng, 2015), which gave an algorithm that is time-efficient (while ours is space-efficient) and randomized (while ours is deterministic) for the case of even $r$ (while ours works for all $r$). Along the way, we provide some new results that generalize technical machinery and yield improvements over previous work. First, we obtain a nearly linear-time randomized algorithm for computing a spectral approximation to the normalized Laplacian for odd $r$. Second, we define and analyze a generalization of the derandomized square for irregular graphs and for sparsifying the product of two distinct graphs. As part of this generalization, we also give a strongly explicit construction of expander graphs of every size.

cs.CC↗

arXiv2019

Faster Matroid Intersection

In this paper we consider the classic matroid intersection problem: given two matroids $\M_{1}=(V,\I_{1})$ and $\M_{2}=(V,\I_{2})$ defined over a common ground set $V$, compute a set $S\in\I_{1}\cap\I_{2}$ of largest possible cardinality, denoted by $r$. We consider this problem both in the setting where each $\M_{i}$ is accessed through an independence oracle, i.e. a routine which returns whether or not a set $S\in\I_{i}$ in $\indep$ time, and the setting where each $\M_{i}$ is accessed through a rank oracle, i.e. a routine which returns the size of the largest independent subset of $S$ in $\M_{i}$ in $\rank$ time. In each setting we provide faster exact and approximate algorithms. Given an independence oracle, we provide an exact $O(nr\log r \indep)$ time algorithm. This improves upon the running time of $O(nr^{1.5} \indep)$ due to Cunningham in 1986 and $\tilde{O}(n^{2} \indep+n^{3})$ due to Lee, Sidford, and Wong in 2015. We also provide two algorithms which compute a $(1-ε)$-approximate solution to matroid intersection running in times $\tilde{O}(n^{1.5}/\eps^{1.5} \indep)$ and $\tilde{O}((n^{2}r^{-1}ε^{-2}+r^{1.5}ε^{-4.5}) \indep)$, respectively. These results improve upon the $O(nr/\eps \indep)$-time algorithm of Cunningham as noted recently by Chekuri and Quanrud. Given a rank oracle, we provide algorithms with even better dependence on $n$ and $r$. We provide an $O(n\sqrt{r}\log n \rank)$-time exact algorithm and an $O(nε^{-1}\log n \rank)$-time algorithm which obtains a $(1-\eps)$-approximation to the matroid intersection problem. The former result improves over the $\tilde{O}(nr \rankt+n^{3})$-time algorithm by Lee, Sidford, and Wong. The rank oracle is of particular interest as the matroid intersection problem with this oracle is a special case of the submodular function minimization problem with an evaluation oracle.

cs.DS↗

arXiv2019

Faster Energy Maximization for Faster Maximum Flow

In this paper we provide an algorithm which given any $m$-edge $n$-vertex directed graph with integer capacities at most $U$ computes a maximum $s$-$t$ flow for any vertices $s$ and $t$ in $m^{11/8+o(1)}U^{1/4}$ time with high probability. This running time improves upon the previous best of $\tilde{O}(m^{10/7} U^{1/7})$ (Mądry 2016), $\tilde{O}(m \sqrt{n} \log U)$ (Lee Sidford 2014), and $O(mn)$ (Orlin 2013) when the graph is not too dense or has large capacities. We achieve this result by leveraging recent advances in solving undirected flow problems on graphs. We show that in the maximum flow framework of (Mądry 2016) the problem of optimizing the amount of perturbation of the central path needed to maximize energy and thereby reduce congestion can be efficiently reduced to a smoothed $\ell_2$-$\ell_p$ flow optimization problem, which can be solved approximately via recent work (Kyng, Peng, Sachdeva, Wang 2019). Leveraging this new primitive, we provide a new long-step interior point method for maximum flow with faster convergence and simpler analysis that no longer needs global potential functions involving energy as in previous methods (Mądry 2013, Mądry 2016).

cs.DS↗

arXiv2019

Principal Component Projection and Regression in Nearly Linear Time through Asymmetric SVRG

Given a data matrix $\mathbf{A} \in \mathbb{R}^{n \times d}$, principal component projection (PCP) and principal component regression (PCR), i.e. projection and regression restricted to the top-eigenspace of $\mathbf{A}$, are fundamental problems in machine learning, optimization, and numerical analysis. In this paper we provide the first algorithms that solve these problems in nearly linear time for fixed eigenvalue distribution and large n. This improves upon previous methods which have superlinear running times when both the number of top eigenvalues and inverse gap between eigenspaces is large. We achieve our results by applying rational approximations to reduce PCP and PCR to solving asymmetric linear systems which we solve by a variant of SVRG. We corroborate these findings with preliminary empirical experiments.

cs.DS↗