$K$-Best Solutions of MSO Problems on Tree-Decomposable Graphs
Data Structures and Algorithms
2017-03-09 v1
Abstract
We show that, for any graph optimization problem in which the feasible solutions can be expressed by a formula in monadic second-order logic describing sets of vertices or edges and in which the goal is to minimize the sum of the weights in the selected sets, we can find the best solutions for -vertex graphs of bounded treewidth in time . In particular, this applies to the problem of finding the shortest simple paths between given vertices in directed graphs of bounded treewidth, giving an exponential speedup in the per-path cost over previous algorithms.
Cite
@article{arxiv.1703.02784,
title = {$K$-Best Solutions of MSO Problems on Tree-Decomposable Graphs},
author = {David Eppstein and Denis Kurz},
journal= {arXiv preprint arXiv:1703.02784},
year = {2017}
}
Comments
14 pages, 0 figures, submitted to the 44th International Colloquium on Automata, Languages, and Programming (ICALP 2017)