首页/论文/1908.00351

arXiv 1908.003512019-08cs.DScs.CG

Sparse Regression via Range Counting

Jean Cardinal, Aurélien Ooms

英文摘要：

The sparse regression problem, also known as best subset selection problem, can be cast as follows: Given a set $S$ of $n$ points in $\mathbb{R}^d$, a point $y\in \mathbb{R}^d$, and an integer $2 \leq k \leq d$, find an affine combination of at most $k$ points of $S$ that is nearest to $y$. We describe a $O(n^{k-1} \log^{d-k+2} n)$-time randomized $(1+\varepsilon)$-approximation algorithm for this problem with $d$ and $\varepsilon$ constant. This is the first algorithm for this problem running in time $o(n^k)$. Its running time is similar to the query time of a data structure recently proposed by Har-Peled, Indyk, and Mahabadi (ICALP'18), while not requiring any preprocessing. Up to polylogarithmic factors, it matches a conditional lower bound relying on a conjecture about affine degeneracy testing. In the special case where $k = d = O(1)$, we also provide a simple $O_δ(n^{d-1+δ})$-time deterministic exact algorithm, for any $δ> 0$. Finally, we show how to adapt the approximation algorithm for the sparse linear regression and sparse convex regression problems with the same running time, up to polylogarithmic factors.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

阅读原文

阅读 PDF ↗

arXiv 页面 ↗HTML 原文 ↗

本文内容

英文摘要相关论文

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

继续探索

相关论文

稀疏超图中匹配的近似计数

Approximate Counting of Matchings in Sparse Hypergraphs

本文给出了一类稀疏一致超图中所有匹配个数的全多项式随机逼近方案。我们的方法是基于推广的典型路径方法的情况下，一致超图。

2012-04-05cs.DMcs.DS

Finding and Counting Patterns in Sparse Graphs

We consider algorithms for finding and counting small, fixed graphs in sparse host graphs. In the non-sparse setting, the parameters treedepth and treewidth play a crucial role in fast, constant-space and polynomial-space algorithms respectively. We discover two new parameters that we call matched treedepth and matched treewidth. We show that finding and counting patterns with low matched treedepth and low matched treewidth can be done asymptotically faster than the existing algorithms when the host graphs are sparse for many patterns. As an application to finding and counting fixed-size patte…

2023-01-06cs.CCcs.DS

Tiered Sampling：一种高效的海量图流稀疏模体近似计数方法

Tiered Sampling: An Efficient Method for Approximate Counting Sparse Motifs in Massive Graph Streams

我们介绍了分层采样，一种新的技术，用于近似计数稀疏图案在大规模的图形，其边缘观察到的流。我们的技术只需要一个单一的数据通过，并使用固定大小的内存$M$，这可以是幅度小于边缘的数量。我们的方法解决了具有挑战性的任务，计数稀疏图案-子图模式，有低概率出现在一个样本的$M$边缘的图形，这是最大量的数据可用于算法在每一步。为了获得计数的无偏和低方差估计，我们将可用存储器划分为储层样本的层（层）。虽然基础层是边缘的标准储集层样本，但其他层是所需基序的子结构的储集层样本。通过存储主题的更频繁的子结构，我们增加了检测到我们正在计数的稀疏主题的出现的概率，从而降低了估计的方差和误差。我们证明了我们的方法的优势，在具体的应用程序中计数稀疏4和5团在大量的图形。我们提出了一个完整的分析和广泛的实验结果，使用合成和真实世界的数据。我们的研究结果表明，我们的方法在获得高质量的近似的4和5团的大型图形使用非常有限的内存量，显着优于单边样本方法计数稀疏图案在大规模的图形。

2017-10-05cs.DS

非常稀疏压缩计数的稀疏恢复

Sparse Recovery with Very Sparse Compressed Counting

Rutgers University（罗格斯大学）

压缩感知（稀疏信号恢复）经常遇到非负数据（例如，图像）。最近，我们开发了使用（密集）压缩计数恢复非负K-稀疏信号的方法。在本文中，我们采用非常稀疏的压缩计数的非负信号恢复。我们的设计矩阵是从最大偏态p-稳定分布（0<p<1）中采样的，我们对设计矩阵进行了稀疏化，使得平均（1-g）-分数项变为零。这个想法与非常稀疏的稳定随机投影（Li et al 2006 and Li 2007）有关，这是估计数据汇总统计量的先前工作。在我们的理论分析中，我们表明，当p->0时，使用M= K/（1-exp（-gK）log N测量就足够了，使得可以在坐标的一次扫描中恢复所有坐标。如果g = 1（即，密集设计），则M = K log N。如果g= 1/K或2/K（即，非常稀疏的设计），则M = 1.58K log N或M = 1.16K log N。这意味着设计矩阵实际上可以非常稀疏，样本复杂性只有很小的膨胀。有趣的是，当p->1时，所需的测量次数基本上是M = 2.7K log N，假设g= 1/K。事实证明，这个结果是一个一般的最坏情况界限。

2013-12-31cs.DScs.ITcs.LG

Accelerating Clique Counting in Sparse Real-World Graphs via Communication-Reducing Optimizations

Counting instances of specific subgraphs in a larger graph is an important problem in graph mining. Finding cliques of size k (k-cliques) is one example of this NP-hard problem. Different algorithms for clique counting avoid counting the same clique multiple times by pivoting or ordering the graph. Ordering-based algorithms include an ordering step to direct the edges in the input graph, and a counting step, which is dominated by building node or edge-induced subgraphs. Of the ordering-based algorithms, kClist is the state-of-the art algorithm designed to work on sparse real-world graphs. Desp…

2021-12-29cs.DCcs.DScs.PF

一种用于稀疏图中子图计数的动态数据结构

A dynamic data structure for counting subgraphs in sparse graphs

我们提出了一个表示图G的动态数据结构，它允许从图G中添加和删除边，并且可以确定一个有界大小的图作为G的导出子图的出现次数.查询在恒定的时间内得到回答。当该数据结构用于表示来自具有有界扩展的类的图时（其包括平面图和更一般地在拓扑子式上闭合的所有真类，以及具有有界平均度的许多其他自然类的图），更新的摊销时间复杂度是多对数的。

2013-01-03cs.DMcs.DSmath.CO

更多相关论文收起更多论文

稀疏一致超图中匹配的近似计数

Approximate Counting of Matchings in Sparse Uniform Hypergraphs

本文给出了交图含爪少的k-一致超图的匹配数的一个全多项式随机逼近方案。我们的方法给出了一个推广的标准路径方法的Jerrum和Sinclair的超图满足局部限制。我们的证明方法依赖于应用欧拉巡回技术的规范路径的基础马尔可夫链。另一方面，我们证明了在没有上述限制的情况下，即使对于k-一致、2-正则和线性超图类，对所有k >= 6，也是NP-难的。

2012-04-24cs.DMcs.DS

Hardness and Algorithms for Robust and Sparse Optimization

The University of Texas at Austin（德克萨斯大学奥斯汀分校）；Massachusetts Institute of Technology（麻省理工学院）；Carnegie Mellon University（卡内基梅隆大学）

We explore algorithms and limitations for sparse optimization problems such as sparse linear regression and robust linear regression. The goal of the sparse linear regression problem is to identify a small number of key features, while the goal of the robust linear regression problem is to identify a small number of erroneous measurements. Specifically, the sparse linear regression problem seeks a $k$-sparse vector $x\in\mathbb{R}^d$ to minimize $\|Ax-b\|_2$, given an input matrix $A\in\mathbb{R}^{n\times d}$ and a target vector $b\in\mathbb{R}^n$, while the robust linear regression problem se…

2022-06-29cs.DScs.LG

近似模型计数、稀疏 XOR 约束与最小距离

Approximate Model Counting, Sparse XOR Constraints and Minimum Distance

本文研究近似模型计数中短稀疏 XOR 约束的可用界限，证明其与模型集几何及最小 Hamming 距离相关，并通过实验讨论改进方向。

2019-07-11cs.DScs.LO

Sparse Hashing for Scalable Approximate Model Counting: Theory and Practice

National University of Singapore（新加坡国立大学）；Indian Institute of Technology, Bombay（印度理工学院孟买分校）

Given a CNF formula F on n variables, the problem of model counting or #SAT is to compute the number of satisfying assignments of F . Model counting is a fundamental but hard problem in computer science with varied applications. Recent years have witnessed a surge of effort towards developing efficient algorithmic techniques that combine the classical 2-universal hashing with the remarkable progress in SAT solving over the past decade. These techniques augment the CNF formula F with random XOR constraints and invoke an NP oracle repeatedly on the resultant CNF-XOR formulas. In practice, calls …

2020-04-30cs.DScs.LGcs.LO

Compressed Sensing with Adversarial Sparse Noise via L1 Regression

UT Austin（德克萨斯大学奥斯汀分校）

We present a simple and effective algorithm for the problem of \emph{sparse robust linear regression}. In this problem, one would like to estimate a sparse vector $w^* \in \mathbb{R}^n$ from linear measurements corrupted by sparse noise that can arbitrarily change an adversarially chosen $η$ fraction of measured responses $y$, as well as introduce bounded norm noise to the responses. For Gaussian measurements, we show that a simple algorithm based on L1 regression can successfully estimate $w^*$ for any $η< η_0 \approx 0.239$, and that this threshold is tight for the algorithm. The number of m…

2019-01-06cs.DScs.LG

Data-compression for Parametrized Counting Problems on Sparse graphs

We study the concept of \emph{compactor}, which may be seen as a counting-analogue of kernelization in counting parameterized complexity. For a function $F:Σ^*\to \Bbb{N}$ and a parameterization $κ: Σ^*\to \Bbb{N}$, a compactor $({\sf P},{\sf M})$ consists of a polynomial-time computable function ${\sf P}$, called \emph{condenser}, and a computable function ${\sf M}$, called \emph{extractor}, such that $F={\sf M}\circ {\sf P}$, and the condensing ${\sf P}(x)$ of $x$ has length at most $s(κ(x))$, for any input $x\in Σ^*.$ If $s$ is a polynomial function, then the compactor is said to be of poly…

2018-09-25cs.DSmath.CO

条件稀疏线性回归

Conditional Sparse Linear Regression

Washington University in St. Louis（圣路易斯华盛顿大学）

机器学习和统计学通常专注于构建捕获绝大多数数据的模型，可能会忽略一小部分数据作为“噪音”或“离群值”。“相比之下，在这里我们考虑联合识别一个重要的（但可能很小的）群体部分的问题，其中有一个高度稀疏的线性回归拟合，以及线性拟合的系数。我们认为，这些任务是有趣的，因为模型本身可能能够在这种特殊情况下实现更好的预测，但也因为它们可以帮助我们理解的数据。我们给这样的问题的算法下的超级规范，当这个未知的部分人口描述的k-DNF条件和回归拟合是s-稀疏常数k和s。对于回归拟合不太稀疏或使用期望误差时该问题的变体，我们也给出了初步的算法，并强调该问题是未来工作的挑战。

2016-08-18cs.DScs.LGstat.ML