We consider approximations for computing minimum weighted cuts in directed graphs. We consider both rooted and global minimum cuts, and both edge-cuts and vertex-cuts. For these problems we give randomized Monte Carlo algorithms that compute a $(1+\epsilon)$-approximate minimum cut in $\tilde{O}(n^2 / \epsilon^2)$ time. These results extend and build on recent work [4] that obtained exact algorithms with similar running times in directed graphs with small integer capacities.
翻译:我们考虑在定向图表中计算最低加权削减的近似值。 我们考虑的是根值和全球最低削减值, 以及边缘切口和顶端切口。 对于这些问题,我们给出了随机的蒙特卡洛算法, 计算出$( 1 ⁇ epsilon)$- 近似最低削减值, 以 $\ tilde{O} (n ⁇ 2 /\ epsilon2) 计值。 这些结果扩展并借鉴了最近的工作 [4], 后者在有小整数的定向图表中获得了类似运行时间的精确算法 。