AI generates interpretation videos

AI extracts and analyses the key points of the paper to generate videos automatically Go Generating

AI Traceability

AI parses the academic lineage of this thesis Generate MRT

AI Insight

AI extracts a summary of this paper

Weibo:
This algorithm is optimal up to o(1) factors assuming recent hardness conjectures: we show by a straightforward reduction that k-Cut on even a simple graph is as hard as -clique, establishing a lower bound of n(1−o(1))k for k-Cut

Faster Minimum k-cut of a Simple Graph

2019 IEEE 60TH ANNUAL SYMPOSIUM ON FOUNDATIONS OF COMPUTER SCIENCE (FOCS 2019), pp.1056.0-1077, (2019)

Cited by: 8|Views250
EI
Full Text
Bibtex
Weibo

Abstract

We consider the (exact, minimum) k-CUT problem: given a graph and an integer k, delete a minimum-weight set of edges so that the remaining graph has at least k connected components. This problem is a natural generalization of the global minimum cut problem, where the goal is to break the graph into k = 2 pieces. Our main result is a (comb...More

Code:

Data:

0
Introduction
• The authors consider the k-Cut problem: given a graph and an integer k, delete a minimumweight set of edges so that the remaining graph has at least k connected components.
• The authors apply a reduction in  which, at a multiplicative cost of O(nk) in the running time, produces a tree T such that exactly k − 1 edges of T have endpoints in different components of the minimum k-cut.
• Since s–t minimum cut is polynomial time solvable, the algorithm can detect which edges to contract.
Highlights
• We consider the k-Cut problem: given a graph and an integer k, delete a minimumweight set of edges so that the remaining graph has at least k connected components. This problem is a natural generalization of the global minimum cut problem, where the goal is to break the graph into k = 2 pieces
• This problem has been actively studied in theory of both exact and approximation algorithms, where each result brought new insights and tools on graph cut algorithms
• The fastest algorithm for general edge weights is due to Chekuri et al 
Results
• For each connected component C in H, there exists a spanning tree TC of C satisfying the following property: Let U be the set of endpoints of edges in TC (more formally, U := (u,v)∈E(TC){u, v}).
• Observe that if Conditions 4.5 and 4.6 hold, for each connected component Ci∗, there exists a set Ui∗ ∈ U of size |V (Ci∗)| such that each vertex vj∗ in Ci∗ belongs on a branch T (u) (u ∈ Ui∗).
• After running Algorithm 4.7, Conditions 4.5 and 4.6 hold, and for each connected component Ci∗, there exists a set Ui∗ ∈ U of size |V (Ci∗)| such that each vertex vj∗ in Ci∗ belongs on a branch T (u) (u ∈ U ).
• The ancestor cut has one edge sharing a maximal branch with each vertex in MinElts(Ui).
• The authors want a set of disjoint branches, one containing each minimal vertex, such that a variant of Lemma 4.4 still holds, so that the authors can set up a similar minimum s–t cut problem.
• For each connected component C in H, there exists a spanning tree TC of C satisfying the following property: Let U be the set of endpoints of edges in TC.
• Construct the graph Gb as follows: the add that edge with endpoints b(u), vertex b(v) in set is i Bi∗, and for each edge in Gb. Note that if the authors contract each
• Fix any component Ci∗, and let U be the set of endpoints in T ′ of edges in Ti∗.
Conclusion
• For each connected component C of H′, let C+ be the set of vertices in H contracted to a vertex in C.
• After running Algorithms 4.25 and 4.30, Conditions 4.23, 4.24, 4.26, 4.27, and 4.28 hold, and for each Ci+, there exists a set Ui+ ∈ U whose subtrees contain precisely all vertices vj∗ in Ci+.
• The minimum ancestor p-cut is the following problem: For each i ∈ [l], remove at least one edge in T (note: not T ′(ui)) such that after removal, no sj ∈ V (T) is in the same component as ui, and such that exactly p − 1 edges are removed in total.
Related work
• The k-Cut problem has been studied extensively in the approximate and fixed-parameter settings as well.

Approximation algorithms. The first approximation algorithm k-Cut was a 2(1−1/k)-approximation of Saran and Vazirani . Later, Naor and Rabani , and also Ravi and Sinha  gave 2-approximation algorithms using tree packing and network strength respectively. Xiao et al  extended Kapoor  and Zhao et al  and generalized Saran and Vazirani to give an (2 − h/k)-approximation in time nO(h). On the hardness front, Manurangsi  showed that for any ǫ > 0, it is NP-hard to achieve a (2 − ǫ)-approximation algorithm in time poly(n, k) assuming the Small Set Expansion Hypothesis.
Reference
Author  