arXiv preprint ·
Faster Cluster Decomposition with Stronger Bi-Criteria Guarantee
- Published
- Source
- arXiv preprint, posted before peer review
- Original
- Read it at arXiv
- Topics
Abstract
Partitioning the vertex set of a graph $G=(V,E)$ into clusters is a classical problem in computer science. As vertices within clusters should be well connected and cuts between clusters sparse, Oveis Gharan and Trevisan (SODA 2014, [OGT14]) proposed to measure the partition quality using inner and outer conductance, i.e., the bi-criteria guarantee. Specifically, when partitioning $V$ into vertex-disjoint subsets $P_1,\ldots, P_k$, every subgraph induced by $P_i$ should have high conductance, while the conductance of $P_i$ in $G$ should be low. They studied the existence of such a partition and designed a polynomial-time algorithm achieving this guarantee. For more than a decade, their algorithm remained the only one achieving this guarantee on its output clusters. In this paper, we study faster cluster decomposition algorithms under the same setup as [OGT14] and prove three results: 1. We make their existential result algorithmic: Our algorithms run in nearly-linear time, and achieve the same bi-criteria guarantee as the one in [OGT14] up to constant factors, including the asymptotically tight dependence on $k$. 2. We prove that their existential result holds algorithmically under a much weaker condition on the initial disjoint subsets. To achieve this, we develop a general algorithm framework that boosts the approximation guarantee of a solution: given an $α$-approximate solution to the $k$-way conductance problem, if $ρ_G(k+1)/ρ_G(k)=Ω\left(α/\varepsilon\right)$, then a $(1+\varepsilon)$-approximate solution can be obtained using only $O(k(1+\logα))$ minimum $s$-$t$ cut computations. 3. We present a nearly-linear time algorithm that finds clusters with the bi-criteria guarantee. Our algorithm has the same guarantee as the one in [OGT14], which, for any unweighted graph of $n$ vertices and $m$ edges, runs in $\widetilde{O}(knm^2)$ time.
The authors’ abstract, as arXiv lists it. arXiv’s listings, abstracts included, are free for anyone to reuse.
Read it on ReadSearch
Each opens after you create an account, and brings you back to this paper.
More in Data Structures & Algorithms
Noncrossing is All You Need: Beyond Identical Delay Functions in Online Matching with Delays on General Metrics
Quadratic-Time Barriers from Self-Referential SAT
Sublinear Metric Distortion for Matching
Exact $\ell_p$-CVP via Euclidean SVP in Polynomial Space
Near linear mixing of random-cluster Glauber dynamics on random regular graphs
Get the few papers worth your time in Data Structures & Algorithms every morning.
ReadSearch reads the new papers in your field each day and brings you the few worth reading, each with its key finding in a sentence.