Download presentation
Presentation is loading. Please wait.
Published byramaic ΠολίΟΞ·Ο Modified over 6 years ago
1
Approximation and Kernelization for Chordal Vertex Deletion
Bart M. P. Jansen & Marcin Pilipczuk SODA 2017 January 18th 2017, Barcelona, Spain
2
Graph modification problems
Make a minimum number of changes to obtain some graph property All NP-complete [Lewis&Yannakakis, JCSSβ80] Vertex Cover Delete vertices to get edgeless graph Feedback Vertex Set Delete vertices to get acyclic graph Vertex Bipartization Delete vertices to get bipartite graph Chordal Vertex Deletion Delete vertices to remove all induced πΆ β₯4
3
Coping with NP-completeness
Polynomial-time approximation algorithms Fixed-parameter tractable algorithms Test for solution of size π in time π π β
π π 1 Kernelization algorithms Reduce to equivalent input of bounded size in polynomial time Function π:βββ is the size of the kernelization Can be exponential or polynomial; smaller is better πΊ π vertices π ππππ¦( πΊ ,π) time πΊβ² π(π) vertices πβ²
4
Coping with graph modification problems
Vertex Cover 2-approximation π β
π π 1 algo 2π-vertex kernel Feedback Vertex Set 2-approximation 3.619 π β
π π 1 algo 2 π 2 +π-vertex kernel Vertex Bipartization π( log π )-approx π β
π π 1 algo π( π 4.5 )-vtx kernel Chordal Vertex Deletion approximation? 2 π π log π β
π π 1 algo polynomial kernel?
5
Difficulty of kernelizing Chordal Vertex Deletion
Graph Class π’ Vertex Deletion is well-understood when: π’ is characterized by finite set of forbidden induced subgraphs Kernelize using sunflower lemma (Vertex Cover) π’ is minor-closed and has bounded treewidth Kernelize using protrusion reduction (Feedback Vertex Set) For chordal graphs: Infinitely many forbidden induced subgraphs πΆ 4 , πΆ 5 , πΆ 6 ,β¦ Not minor closed & unbounded treewidth (cliques are chordal)
6
Our contribution Theorem 1.
Chordal Vertex Del. has a kernel with π π log 58 π vertices Resolves an open problem by Marx from 2006 Required to start the kernelization Theorem 2. Solution of size π( opt 4 log 2 opt) can be found in poly-time After our arXiv preprint in May 2016: bounds have been improved [next talk] analogous results for Distance Hereditary Vertex Deletion [Kim & Kwon, arXivβ16]
7
ErdΓΆs-PΓ³sa property for chordless cycles
Lemma. If πΊβ{π£} is chordal, then in poly-time one can find: an integer π, a set of π chordless cycles that pairwise intersect only in π£, and a set πβπ πΊ β{π£} of size β€12π such that πΊβπ is chordal. π£ Can be used to identify vertices that must belong to any optimal solution
8
Sketch of the approximation algorithm
Reduce to the following special case: Input graph πΊ has partition π πΊ =π΄ β π΅ such that πΊ[π΄] is chordal, πΊ[π΅] is a clique Clique πΊ[π΅] Chordal πΊ[π΄]
9
Sketch of the approximation algorithm
Reduce to the following special case: Input graph πΊ has partition π πΊ =π΄ β π΅ such that πΊ[π΄] is chordal, πΊ[π΅] is a clique Fact. Any chordal graph πΊ has a clique separator π΅ such that each component of πΊβπ΅ has at most π πΊ /2 vertices Every instance of Chordal Vertex Deletion has a balanced separator consisting of a clique plus opt vertices
10
Sketch of the approximation algorithm
Reduce to the following special case: Input graph πΊ has partition π πΊ =π΄ β π΅ such that πΊ[π΄] is chordal, πΊ[π΅] is a clique While πΊ has a component πΆ that is not chordal: find balanced separator for πΆ of form clique + π (opt) vertices add the π (opt) vertices to the approximate solution π 0 add the clique π΅ to a list of cliques π΅ 1 , π΅ 2 ,β¦ Terminates in β=π(opt log π) iterations with chordal πΊ[π΄] π πΊ =π΄βͺ π΅ 1 βͺ π΅ 2 βͺβ¦βͺ π΅ β βͺ π 0 Recursion tree has O(log n) levels, no level of the recusion tree has more than OPT non-chordal components because then there would be more than OPT disjoint chordless cycles => solution is larger than OPT. Every instance of Chordal Vertex Deletion has a balanced separator consisting of a clique plus opt vertices
11
Sketch of the approximation algorithm
Reduce to the following special case: Input graph πΊ has partition π πΊ =π΄ β π΅ such that πΊ[π΄] is chordal, πΊ[π΅] is a clique Build up global solution using algorithm for special case: πΊ[π΄] πΊ[ π΄βͺ π΅ 1 β π 1 ] πΊ[ π΄βͺβ¦ β π 2 ] β― Recursion tree has O(log n) levels, no level of the recusion tree has more than OPT non-chordal components because then there would be more than OPT disjoint chordless cycles => solution is larger than OPT. π΅ 1 π΅ 2 π΅ 3 Factor β=π(opt log π ) overhead in approximation guarantee
12
Sketch of the approximation algorithm
Reduce to the following special case: Input graph πΊ has partition π πΊ =π΄ β π΅ such that πΊ[π΄] is chordal, πΊ[π΅] is a clique Reduce to even more special case by another decomposition Relate special-case ChVD solutions to multicuts in a digraph Apply Guptaβs rounding algorithm for fractional multicuts Round fractional ChVD-solution into integral one [Gupta, SODAβ02]
13
Conclusion First polynomial kernel and poly(opt) approximation for ChVD Based on graph-theoretic insights Open: Poly kernel for Directed Feedback Vertex Set? Planarization? Next talk: improvements to kernel size and approximation factor Theorem 1. Chordal Vertex Del. has a kernel with π π log 58 π vertices Theorem 2. Solution of size π( opt 4 log 2 opt) can be found in poly-time
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.