Lossy kernelization

DOI10.1145/3055399.3055456zbMATH Open1370.68136arXiv1604.04111OpenAlexW2911549278MaRDI QIDQ4977974FDOQ4977974

Authors: Daniel Lokshtanov, Fahad Panolan, M. S. Ramanujan, Saket Saurabh 

Publication date: 17 August 2017

Published in: Proceedings of the 49th Annual ACM SIGACT Symposium on Theory of Computing (Search for Journal in Brave)

Abstract: In this paper we propose a new framework for analyzing the performance of preprocessing algorithms. Our framework builds on the notion of kernelization from parameterized complexity. However, as opposed to the original notion of kernelization, our definitions combine well with approximation algorithms and heuristics. The key new definition is that of a polynomial size $alpha$-approximate kernel. Loosely speaking, a polynomial size $alpha$-approximate kernel is a polynomial time pre-processing algorithm that takes as input an instance $(I,k)$ to a parameterized problem, and outputs another instance $(I^{\prime },k^{\prime })$ to the same problem, such that $|I^{\prime }|+k^{\prime }leq{k}^{O(1)}$. Additionally, for every $cgeq1$, a $c$-approximate solution $s^{\prime }$ to the pre-processed instance $(I^{\prime },k^{\prime })$ can be turned in polynomial time into a $(ccdotalpha)$-approximate solution $s$ to the original instance $(I,k)$. Our main technical contribution are $alpha$-approximate kernels of polynomial size for three problems, namely Connected Vertex Cover, Disjoint Cycle Packing and Disjoint Factors. These problems are known not to admit any polynomial size kernels unless $NPsubseteqcoNP/poly$. Our approximate kernels simultaneously beat both the lower bounds on the (normal) kernel size, and the hardness of approximation lower bounds for all three problems. On the negative side we prove that Longest Path parameterized by the length of the path and Set Cover parameterized by the universe size do not admit even an $alpha$-approximate kernel of polynomial size, for any $alphageq1$, unless $NPsubseteqcoNP/poly$. In order to prove this lower bound we need to combine in a non-trivial way the techniques used for showing kernelization lower bounds with the methods for showing hardness of approximation

Full work available at URL: https://arxiv.org/abs/1604.04111

Recommendations

• Kernelization -- preprocessing with a guarantee
• On the approximate compressibility of connected vertex cover
• Lower bounds on kernelization
• Kernelization lower bounds through colors and IDs
• Lossy Kernels for Hitting Subgraphs

zbMATH Keywords

inapproximabilityparameterized complexitykernelization

Mathematics Subject Classification ID

Analysis of algorithms and problem complexity (68Q25) Computational difficulty of problems (lower bounds, completeness, difficulty of approximation, etc.) (68Q17) Approximation algorithms (68W25)

Cited In (51)

• Dynamic kernels for hitting sets and set packing
• \(p\)-edge/vertex-connected vertex cover: parameterized and approximation algorithms
• Packing cycles faster than Erdős-Pósa
• Matroid-constrained vertex cover
• Lossy kernels for connected dominating set on sparse graphs
• On MAX-SAT with cardinality constraint
• Parameterized approximation algorithms for bidirected Steiner network problems
• On the Parameterized Approximability of Contraction to Classes of Chordal Graphs
• To close is easier than to open: dual parameterization to \(k\)-median
• On the kernelization complexity of string problems
• On the approximate compressibility of connected vertex cover
• The parameterized hardness of the \(k\)-center problem in transportation networks
• On the lossy kernelization for connected treedepth deletion set
• Streaming kernelization
• Refined notions of parameterized enumeration kernels with applications to matching cut enumeration
• Approximate Turing Kernelization for Problems Parameterized by Treewidth
• Lossy kernels for graph contraction problems
• Lossy kernels for connected dominating set on sparse graphs
• Revisiting connected vertex cover: FPT algorithms and lossy kernels
• On the parameterized complexity of contraction to generalization of trees
• On data reduction for dynamic vector bin packing
• Search-space reduction via essential vertices
• Parameterized approximation schemes for Steiner trees with small number of Steiner vertices
• On approximate data reduction for the Rural Postman Problem: Theory and experiments
• Safe approximation and its relation to kernelization
• Title not available (Why is that?)
• Data reductions and combinatorial bounds for improved approximation algorithms
• Parameterized complexity of geometric covering problems having conflicts
• The parameterized complexity of cycle packing: indifference is not an issue
• On the parameterized complexity of maximum degree contraction problem
• Time-approximation trade-offs for inapproximable problems
• Hans Bodlaender and the Theory of Kernelization Lower Bounds
• Parameterized Approximation Schemes for Independent Set of Rectangles and Geometric Knapsack
• On MAX-SAT with cardinality constraint
• Parameterized approximation via fidelity preserving transformations
• Lossy planarization: a constant-factor approximate kernelization for planar vertex deletion
• Parameterized approximation via fidelity preserving transformations
• Approximation algorithms inspired by kernelization methods
• Fractals for kernelization lower bounds
• Improved FPT approximation scheme and approximate kernel for biclique-free max \(k\)-weight SAT: greedy strikes back
• Introducing \textsf{lop}-kernels: a framework for kernelization lower bounds
• Fixed-parameter algorithms for unsplittable flow cover
• Parameterized algorithms and data reduction for the short secluded s‐t‐path problem
• Lossy Kernels for Hitting Subgraphs
• Parameterized approximation schemes for Steiner trees with small number of Steiner vertices
• The parameterized hardness of the \(k\)-center problem in transportation networks
• On approximate preprocessing for domination and hitting subgraphs with connected deletion sets
• Partial vertex cover on graphs of bounded degeneracy
• Lossy kernelization of same-size clustering
• Lossy kernelization of same-size clustering
• Packing arc-disjoint cycles in oriented graphs

This page was built for publication: Lossy kernelization

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4977974)