Abstract
We consider the (precedence constrained) Minimum Feedback Arc Set problem with triangle inequalities on the weights, which finds important applications in problems of ranking with inconsistent information. We present a surprising structural insight showing that the problem is a special case of the minimum vertex cover in hypergraphs with edges of size at most 3.
Similar content being viewed by others
Notes
Recall that a 0∖1 solution δ ∗ is minimal if the removal of any arc (i,j) from its support makes it unfeasible.
When (i,j′)∈M v this is the only possible case, namely case (ii) does not hold. Indeed, see Fig. 6, when (i,j′)∈M v , the following is a valid constraint δ v,j′+δ j,k′+δ k,v ≥1 and we are considering the case with δ v,j′=δ j,k′=0.
References
Aharoni, R., Holzman, R., Krivelevich, M.: On a theorem of Lovász on covers in τ-partite hypergraphs. Combinatorica 16(2), 149–174 (1996)
Ailon, N.: Aggregation of partial rankings, p-ratings and top-m lists. Algorithmica 57(2), 284–300 (2010)
Ailon, N., Avigdor-Elgrabli, N., Liberty, E., van Zuylen, A.: Improved approximation algorithms for bipartite correlation clustering. SIAM J. Comput. 41(5), 1110–1121 (2012)
Ailon, N., Charikar, M., Newman, A.: Aggregating inconsistent information: Ranking and clustering. J. ACM 55(5) (2008)
Ambühl, C., Mastrolilli, M.: Single machine precedence constrained scheduling is a vertex cover problem. Algorithmica 53(4), 488–503 (2009)
Ambühl, C., Mastrolilli, M., Mutsanas, N., Svensson, O.: Precedence constraint scheduling and connections to dimension theory of partial orders. Bull. Eur. Assoc. Theor. Comput. Sci. 95, 45–58 (2008)
Ambühl, C., Mastrolilli, M., Mutsanas, N., Svensson, O.: On the approximability of single-machine scheduling with precedence constraints. Math. Oper. Res. 36(4), 653–669 (2011)
Ambühl, C., Mastrolilli, M., Svensson, O.: Inapproximability results for sparsest cut, optimal linear arrangement, and precedence constraint scheduling. In: Proceedings of FOCS, pp. 329–337 (2007)
Bansal, N., Khot, S.: Optimal Long-Code test with one free bit. In: Proceedings of FOCS, pp. 453–462 (2009)
Correa, J.R., Schulz, A.S.: Single machine scheduling with precedence constraints. Math. Oper. Res. 30(4), 1005–1021 (2005)
Even, G., Naor, J., Rao, S., Schieber, B.: Divide-and-conquer approximation algorithms via spreading metrics. J. ACM 47(4), 585–616 (2000)
Even, G., Naor, J., Schieber, B., Sudan, M.: Approximating minimum feedback sets and multicuts in directed graphs. Algorithmica 20(2), 151–174 (1998)
Grötschel, M., Jünger, M., Reinelt, G.: Acyclic subdigraphs and linear orderings: polytopes, facets, and a cutting plane algorithm. In: Graphs and Orders, pp. 217–264 (1985)
Guruswami, V., Håstad, J., Manokaran, R., Raghavendra, P., Charikar, M.: Beating the random ordering is hard: every ordering csp is approximation resistant. SIAM J. Comput. 40(3), 878–914 (2011)
Hall, L.A., Schulz, A.S., Shmoys, D.B., Wein, J.: Scheduling to minimize average completion time: off-line and on-line algorithms. Math. Oper. Res. 22, 513–544 (1997)
Kann, V.: On the Approximability of NP-Complete Optimization Problems. Ph.D. thesis, Department of Numerical Analysis and Computing Science, Royal Institute of Technology, Stockholm (1992)
Karp, R.: Reducibility Among Combinatorial Problems, pp. 85–103. Plenum, New York (1972)
Kenyon-Mathieu, C., Schudy, W.: How to rank with few errors. In: Proceedings of STOC, pp. 95–103 (2007)
Krivelevich, M.: Approximate set covering in uniform hypergraphs. J. Algorithms 25(1), 118–143 (1997)
Kuhn, F., Mastrolilli, M.: Vertex cover in graphs with locally few colors. Inf. Comput. 222, 265–277 (2013)
Lawler, E.L., Lenstra, J.K., Rinnooy Kan, A.H.G., Shmoys, D.B.: Sequencing and scheduling: Algorithms and complexity. In: Graves, S.C., Rinnooy Kan, A.H.G., Zipkin, P. (eds.) Handbooks in Operations Research and Management Science, vol. 4, pp. 445–552. North-Holland, Amsterdam (1993)
Lempel, A., Cederbaum, I.: Minimum feedback arc and vertex sets of a directed graph. IEEE Trans. Circuit Theory 4(13), 399–403 (1966)
Pardalos, P., Du, D.: Handbook of Combinatorial Optimization: Supplement vol. 1. Springer, Berlin (1999)
Schuurman, P., Woeginger, G.J.: Polynomial time approximation algorithms for machine scheduling: ten open problems. J. Sched. 2(5), 203–213 (1999)
Seymour, P.D.: Packing directed circuits fractionally. Combinatorica 15(2), 281–288 (1995)
Trotter, W.T.: Combinatorics and Partially Ordered Sets: Dimension Theory. Johns Hopkins Series in the Mathematical Sciences. Johns Hopkins University Press, Baltimore (1992)
van Zuylen, A., Hegde, R., Jain, K., Williamson, D.P.: Deterministic pivoting algorithms for constrained ranking and clustering problems. In: Proceedings of SODA, pp. 405–414 (2007)
van Zuylen, A., Williamson, D.P.: Deterministic pivoting algorithms for constrained ranking and clustering problems. Math. Oper. Res. 34(3), 594–620 (2009)
Vazirani, V.V.: Approximation Algorithms. Springer, Berlin (2001)
Acknowledgements
To my beloved mom Elsa.
Author information
Authors and Affiliations
Corresponding author
Additional information
This research is supported by Swiss National Science Foundation project N. 200020-144491/1 “Approximation Algorithms for Machine Scheduling Through Theory and Experiments” and by Hasler Foundation Grant 11099.
Appendices
Appendix A: Ranking with Probability Inequalities: a Counterexample
The following example shows that probabilities inequalities are not sufficient for (3a)–(3d) to be a proper formulation:
Consider the instance with 8 nodes with weight zero on the arcs displayed in Fig. 7 (therefore the reversed arcs have weight 1). Moreover, all the arcs in {2,3}×{7,8} have weight 1 (the reversed zero). Finally, all the remaining arcs have weight 0.5, namely those in {1}×{4,5,6} and the reversed ones. A feasible solution for (2a)–(2d) is obtained by picking all the displayed arcs in Fig. 7 and none of the reversed ones (therefore we have to pick also those in {2,3}×{7,8}, {7,8}×{2,3}, {4,5,6}×{1} and {1}×{4,5,6} in order to satisfy the constraints in (2a)–(2d)). This solution has value 7, whereas any total ordering has value not smaller than 7.5 (the best total ordering is (2,3,4,5,6,7,8,1)).
Appendix B: A Comment on Formulation (3a)–(3d)
If the poset is not empty the additional constraints that are present in formulation (3a)–(3d) but not in (2a)–(2d) are also necessary. Indeed, in Figure 8 any permutation that complies with the precedence constraints has value larger than the solution suggested in the picture with a cycle.
Rights and permissions
About this article
Cite this article
Mastrolilli, M. The Feedback Arc Set Problem with Triangle Inequality Is a Vertex Cover Problem. Algorithmica 70, 326–339 (2014). https://doi.org/10.1007/s00453-013-9811-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00453-013-9811-2