An O(Nm) Time Algorithm for Finding the Min Length Directed Cycle in a Graph
Name
Time-Algorithm-James-Orlin.pdf
Size
1.9 MB
Format
Adobe PDF
Checksum (MD5)
9acb88448b2fe773aa3e17512fe73dbd
Author(s)
Sedeno-Noda, Antonio
Orlin, James B
Date Issued
October 2016
Journal
Proceeding SODA '17 Proceedings of the Twenty-Eighth Annual ACM-SIAM Symposium on Discrete Algorithms
Publisher
Society for Industrial and Applied Mathematics
Citation
Orlin, James B. and Antonio Sede ̃no-Noda. "An O(nm) time algorithm for finding the min length directed cycle in a graph." In Proceeding SODA '17 Proceedings of the Twenty-Eighth Annual ACM-SIAM Symposium on Discrete Algorithms, Barcelona, Spain, January 16-19 2017, pp. 1866-1879.
Version
Author's final manuscript
Abstract
In this paper, we introduce an O(nm) time algorithm to determine the minimum length directed cycle (also called the "minimum weight directed cycle") in a directed network with n nodes and m arcs and with no negative length directed cycles. This result improves upon the previous best time bound of O(nm + n² og log n). Our algorithm first determines the cycle with minimum mean length λ* in O(nm) time. Subsequently, it chooses node potentials so that all reduced costs are λ* or greater. It then solves the all pairs shortest path problem, but restricts attention to paths of length at most nλ*. We speed up the shortest path calculations to O(m) per source node, leading to an O(nm) running time in total. We also carry out computational experiments comparing the performance of the proposed methods and other state-of-the-art methods. Experiments confirmed that it is advantageous to solve the minimum mean cycle problem prior to solving shortest path problems. Analysis of our experiments suggest that the running time to solve the minimum length directed cycle problem was much faster than O(n² on average.
MIT Department
Sloan School of Management
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
http://dl.acm.org/citation.cfm?id=3039808
