arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.08032cs.DS
收藏

高效地对多个边缘进行采样

Sampling Multiple Edges Efficiently

Talya Eden, Saleet Mossel, Ronitt Rubinfeld

摘要:

我们提出了一个次线性时间算法,允许一个采样多个边缘的分布是逐点$ε$ -接近均匀分布,在\emph{amortized-efficient}的方式。我们考虑的邻接表查询模型,其中访问图$G$是通过度和邻居查询。Eden和Rosenbaum(SOSA 18)提出了在该模型中对单个边缘进行采样的问题。令$n$和$m$分别表示$G$的顶点和边的数目。Eden和Rosenbaum提供了$Θ^*(n/\sqrt m)$的上界和下界,用于对一般图中的单个边进行采样(其中$O^*(\cdot)$抑制$\textrm{poly}(1/ε)$和$\textrm{poly}(\log n)$依赖性)。我们问是否可以规避查询复杂性下限采样一个单一的边缘时,需要多个样本。也就是说,如果我们允许预处理阶段,我们是否可以获得改进的摊销每个样本成本?我们的回答是肯定的。我们提出了一种算法,如果一个人知道所需的样本$q$的数量提前,有一个整体的成本是次线性的$q$,即$O^*(\sqrt q \cdot(n/\sqrt m))$,这是严格优选的$O^*(q\cdot (n/\sqrt m))$成本从$q$调用的算法由伊登和罗森鲍姆。在这项工作的一个初步版本之后,Tsuntek和Thorup(arXiv,预印本)证明了这个界限本质上是最优的。

英文摘要:

We present a sublinear time algorithm that allows one to sample multiple edges from a distribution that is pointwise $ε$-close to the uniform distribution, in an \emph{amortized-efficient} fashion. We consider the adjacency list query model, where access to a graph $G$ is given via degree and neighbor queries. The problem of sampling a single edge in this model has been raised by Eden and Rosenbaum (SOSA 18). Let $n$ and $m$ denote the number of vertices and edges of $G$, respectively. Eden and Rosenbaum provided upper and lower bounds of $Θ^*(n/\sqrt m)$ for sampling a single edge in general graphs (where $O^*(\cdot)$ suppresses $\textrm{poly}(1/ε)$ and $\textrm{poly}(\log n)$ dependencies). We ask whether the query complexity lower bound for sampling a single edge can be circumvented when multiple samples are required. That is, can we get an improved amortized per-sample cost if we allow a preprocessing phase? We answer in the affirmative. We present an algorithm that, if one knows the number of required samples $q$ in advance, has an overall cost that is sublinear in $q$, namely, $O^*(\sqrt q \cdot(n/\sqrt m))$, which is strictly preferable to $O^*(q\cdot (n/\sqrt m))$ cost resulting from $q$ invocations of the algorithm by Eden and Rosenbaum. Subsequent to a preliminary version of this work, Tětek and Thorup (arXiv, preprint) proved that this bound is essentially optimal.

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。

↑