Algorithms for Discrepancy, Matchings, and Approximations: Fast, Simple, and Practical

• URL: http://arxiv.org/abs/2209.01147v1
• Date: Fri, 2 Sep 2022 15:59:09 GMT
• Title: Algorithms for Discrepancy, Matchings, and Approximations: Fast, Simple, and Practical
• Authors: M\'onika Csik\'os and Nabil H. Mustafa
• Abstract summary: Given a finite set system $(X,mathcal S)$, the emphdiscrepancy of a two-coloring $chi:Xto-1,1$ is defined as $max_S in mathcal S|chi(S)|$. We propose a randomized algorithm which, for any $d>0$ and $(X,mathcal S)$ with dual shatter function $pi*(k)=O(kd)$, returns a coloring with expected discrepancy
• Score: 1.2183405753834562
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We study one of the key tools in data approximation and optimization: low-discrepancy colorings. Formally, given a finite set system $(X,\mathcal S)$, the \emph{discrepancy} of a two-coloring $\chi:X\to\{-1,1\}$ is defined as $\max_{S \in \mathcal S}|{\chi(S)}|$, where $\chi(S)=\sum\limits_{x \in S}\chi(x)$. We propose a randomized algorithm which, for any $d>0$ and $(X,\mathcal S)$ with dual shatter function $\pi^*(k)=O(k^d)$, returns a coloring with expected discrepancy $O\left({\sqrt{|X|^{1-1/d}\log|\mathcal S|}}\right)$ (this bound is tight) in time $\tilde O\left({|\mathcal S|\cdot|X|^{1/d}+|X|^{2+1/d}}\right)$, improving upon the previous-best time of $O\left(|\mathcal S|\cdot|X|^3\right)$ by at least a factor of $|X|^{2-1/d}$ when $|\mathcal S|\geq|X|$. This setup includes many geometric classes, families of bounded dual VC-dimension, and others. As an immediate consequence, we obtain an improved algorithm to construct $\varepsilon$-approximations of sub-quadratic size. Our method uses primal-dual reweighing with an improved analysis of randomly updated weights and exploits the structural properties of the set system via matchings with low crossing number -- a fundamental structure in computational geometry. In particular, we get the same $|X|^{2-1/d}$ factor speed-up on the construction time of matchings with crossing number $O\left({|X|^{1-1/d}}\right)$, which is the first improvement since the 1980s. The proposed algorithms are very simple, which makes it possible, for the first time, to compute colorings with near-optimal discrepancies and near-optimal sized approximations for abstract and geometric set systems in dimensions higher than $2$.

Related papers

• Complexity of Minimizing Projected-Gradient-Dominated Functions with Stochastic First-order Oracles [38.45952947660789]
  This work investigates the performance limits of projected first-order methods for minimizing functions under the $(alpha,tau,mathcal)$-projected-dominance property.
  arXiv  Detail & Related papers  (2024-08-03T18:34:23Z)
• Efficient Continual Finite-Sum Minimization [52.5238287567572]
  We propose a key twist into the finite-sum minimization, dubbed as continual finite-sum minimization. Our approach significantly improves upon the $mathcalO(n/epsilon)$ FOs that $mathrmStochasticGradientDescent$ requires. We also prove that there is no natural first-order method with $mathcalOleft(n/epsilonalpharight)$ complexity gradient for $alpha 1/4$, establishing that the first-order complexity of our method is nearly tight.
  arXiv  Detail & Related papers  (2024-06-07T08:26:31Z)
• Dueling Optimization with a Monotone Adversary [35.850072415395395]
  We study the problem of dueling optimization with a monotone adversary, which is a generalization of (noiseless) dueling convex optimization. The goal is to design an online algorithm to find a minimizer $mathbfx*$ for a function $fcolon X to mathbbRd.
  arXiv  Detail & Related papers  (2023-11-18T23:55:59Z)
• A Whole New Ball Game: A Primal Accelerated Method for Matrix Games and Minimizing the Maximum of Smooth Functions [44.655316553524855]
  We design algorithms for minimizing $max_iin[n] f_i(x) over a $d$-dimensional Euclidean or simplex domain. When each $f_i$ is $1$-Lipschitz and $1$-smooth, our method computes an $epsilon-approximate solution.
  arXiv  Detail & Related papers  (2023-11-17T22:07:18Z)
• Memory-Constrained Algorithms for Convex Optimization via Recursive Cutting-Planes [23.94542304111204]
  First class of algorithms that provides a positive trade-off between gradient descent and cutting-plane methods in any regime with $epsilonleq 1/sqrt d$. In the regime $epsilon leq d-Omega(d)$, our algorithm with $p=d$ achieves the information-theoretic optimal memory usage and improves the oracle-complexity of gradient descent.
  arXiv  Detail & Related papers  (2023-06-16T17:00:51Z)
• Fast $(1+\varepsilon)$-Approximation Algorithms for Binary Matrix Factorization [54.29685789885059]
  We introduce efficient $(1+varepsilon)$-approximation algorithms for the binary matrix factorization (BMF) problem. The goal is to approximate $mathbfA$ as a product of low-rank factors. Our techniques generalize to other common variants of the BMF problem.
  arXiv  Detail & Related papers  (2023-06-02T18:55:27Z)
• Randomized Block-Coordinate Optimistic Gradient Algorithms for Root-Finding Problems [8.0153031008486]
  We develop two new algorithms to approximate a solution of nonlinear equations in large-scale settings. We apply our methods to a class of large-scale finite-sum inclusions, which covers prominent applications in machine learning.
  arXiv  Detail & Related papers  (2023-01-08T21:46:27Z)
• An Optimal Algorithm for Strongly Convex Min-min Optimization [79.11017157526815]
  Existing optimal first-order methods require $mathcalO(sqrtmaxkappa_x,kappa_y log 1/epsilon)$ of computations of both $nabla_x f(x,y)$ and $nabla_y f(x,y)$. We propose a new algorithm that only requires $mathcalO(sqrtkappa_x log 1/epsilon)$ of computations of $nabla_x f(x,
  arXiv  Detail & Related papers  (2022-12-29T19:26:12Z)
• Sketching Algorithms and Lower Bounds for Ridge Regression [65.0720777731368]
  We give a sketching-based iterative algorithm that computes $1+varepsilon$ approximate solutions for the ridge regression problem. We also show that this algorithm can be used to give faster algorithms for kernel ridge regression.
  arXiv  Detail & Related papers  (2022-04-13T22:18:47Z)
• Optimal Regret Algorithm for Pseudo-1d Bandit Convex Optimization [51.23789922123412]
  We study online learning with bandit feedback (i.e. learner has access to only zeroth-order oracle) where cost/reward functions admit a "pseudo-1d" structure. We show a lower bound of $min(sqrtdT, T3/4)$ for the regret of any algorithm, where $T$ is the number of rounds. We propose a new algorithm sbcalg that combines randomized online gradient descent with a kernelized exponential weights method to exploit the pseudo-1d structure effectively.
  arXiv  Detail & Related papers  (2021-02-15T08:16:51Z)
• On the Complexity of Minimizing Convex Finite Sums Without Using the Indices of the Individual Functions [62.01594253618911]
  We exploit the finite noise structure of finite sums to derive a matching $O(n2)$-upper bound under the global oracle model. Following a similar approach, we propose a novel adaptation of SVRG which is both emphcompatible with oracles, and achieves complexity bounds of $tildeO(n2+nsqrtL/mu)log (1/epsilon)$ and $O(nsqrtL/epsilon)$, for $mu>0$ and $mu=0$
  arXiv  Detail & Related papers  (2020-02-09T03:39:46Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.