arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2305.08353cs.DScs.LG
收藏

具有截止期实例的快速高效匹配算法

Fast and Efficient Matching Algorithm with Deadline Instances

Simons Institute for the Theory of Computing, UC Berkeley(加州大学伯克利分校西蒙斯理论计算研究所);Johns Hopkins University(约翰斯·霍普金斯大学);University of Texas at Austin(德克萨斯大学奥斯汀分校);Rice University(莱斯大学)

Zhao Song, Weixin Wang, Chenbo Yin, Junze Yin

AI总结:

针对带截止期的在线加权匹配,本文提出 FastGreedy 与 FastPostponedGreedy,并用草图矩阵将复杂度降至 $\widetilde{O}(ε^{-2}(n+d))$,竞争比仅损失 $1-ε$。

摘要:

在线加权匹配问题是机器学习中的一个基础问题,因为它有着众多应用。尽管该领域已有许多努力,但现有算法要么太慢,要么没有考虑 $\mathrm{deadline}$(节点可被匹配的最长时间)。本文首先引入一个带有 $\mathrm{deadline}$ 的市场模型。接着,我们提出两个优化算法(\textsc{FastGreedy} 和 \textsc{FastPostponedGreedy}),并给出算法时间复杂度和正确性的理论证明。在 \textsc{FastGreedy} 算法中,我们已经知道一个节点是买方还是卖方;但在 \textsc{FastPostponedGreedy} 算法中,每个节点的身份最初是未知的。然后,我们推广了一种草图矩阵,以便在真实数据集和合成数据集上运行原始算法及我们的算法。令 $ε\in(0,0.1)$ 表示每条边真实权重的相对误差。原始 \textsc{Greedy} 和 \textsc{PostponedGreedy} 的竞争比分别为 $\frac{1}{2}$ 和 $\frac{1}{4}$。在这两个原始算法基础上,我们提出 \textsc{FastGreedy} 和 \textsc{FastPostponedGreedy} 算法,其竞争比分别为 $\frac{1-ε}{2}$ 和 $\frac{1-ε}{4}$。同时,我们的算法比原来两个算法运行更快。给定 $\mathbb{R}^d$ 中的 $n$ 个节点,我们将时间复杂度从 $O(nd)$ 降至 $\widetilde{O}(ε^{-2}\cdot(n+d))$,其中对任意函数 $f$,用 $\widetilde{O}(f)$ 表示 $f\cdot\mathrm{poly}(\log f)$。

英文摘要:

The online weighted matching problem is a fundamental problem in machine learning due to its numerous applications. Despite many efforts in this area, existing algorithms are either too slow or don't take $\mathrm{deadline}$ (the longest time a node can be matched) into account. In this paper, we introduce a market model with $\mathrm{deadline}$ first. Next, we present our two optimized algorithms (\textsc{FastGreedy} and \textsc{FastPostponedGreedy}) and offer theoretical proof of the time complexity and correctness of our algorithms. In \textsc{FastGreedy} algorithm, we have already known if a node is a buyer or a seller. But in \textsc{FastPostponedGreedy} algorithm, the status of each node is unknown at first. Then, we generalize a sketching matrix to run the original and our algorithms on both real data sets and synthetic data sets. Let $ε\in (0,0.1)$ denote the relative error of the real weight of each edge. The competitive ratio of original \textsc{Greedy} and \textsc{PostponedGreedy} is $\frac{1}{2}$ and $\frac{1}{4}$ respectively. Based on these two original algorithms, we proposed \textsc{FastGreedy} and \textsc{FastPostponedGreedy} algorithms and the competitive ratio of them is $\frac{1 - ε}{2}$ and $\frac{1 - ε}{4}$ respectively. At the same time, our algorithms run faster than the original two algorithms. Given $n$ nodes in $\mathbb{R} ^ d$, we decrease the time complexity from $O(nd)$ to $\widetilde{O}(ε^{-2} \cdot (n + d))$, where for any function $f$, we use $\widetilde{O}(f)$ to denote $f \cdot \mathrm{poly}(\log f)$.

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。

↑