Michel X. Goemans
David Williamson
ABSTRACT
A number of recent papers on approximation algorithms have used the square roots of unity, -1 and 1 to represent binary decision variables for problems in combinatorial optimization, and have relaxed these to unit vectors in real space using semidefinite programming in order to obtain near optimal solutions to these problems. In this paper, we consider using the cube roots of unity, 1, ei2π/3, to represent ternary decision variables for problems in combinatorial optimization. Here the natural relaxation is that of unit vectors in complex space. We use an extension of semidefinite programming to complex space to solve the natural relaxation, and use a natural extension of the random hyperplane technique introduced by the authors in [8] to obtain near-optimal solutions to the problems.
In particular, we consider the problem of maximizing the total weight of satisfied equations xu-xv≡c (mod 3) and inequations xu-xv≢c (mod 3) , where xu ∈ {0,1,2} u. This problem can be used to model the MAX-3-CUT problem and a directed variant we call MAX-3-DICUT. For the general problem, we obtain a .79373-approximation algorithm. If the instance contains only inequations (as it does for MAX-3-CUT), we obtain a performance guarantee of 7/12 + 3/(4π2) arccos2 (-1/4)≈.83601. This compares with proven performance guarantees of .800217 for MAX-3-CUT (by Frieze and Jerrum [7]) and 1/3 + 10 -8 for the general problem (by Andersson, Engebretson, and Håstad [2]). It matches the guarantee of .836008 for MAX-3-CUT found independently by de Klerk, Pasechnik, and Warners [4]. We show that all these algorithms are in fact identical in the case of MAX-3-CUT.
- 1.F.Aliz deh.Interior point methods in semide .nite programming with applications to combinatorial optimization.SIAM Journal on Optimization 5:13 -51,1995.Google Scholar
- 2.G.Andersson,L.Engebretsen,and J.Hast d.A new wy to use semide .nite programming with pplic tions to line r equations mod p .Journal of Algorithms 39:162 -204,2001. Google ScholarDigital Library
- 3.B.Chor nd M.Sudan.A geometric approach to betweenness.SIAM Journal on Discrete Mathematics 11:511 -523,1998. Google ScholarDigital Library
- 4.E.deKlerk,D.V.Psechnik,ndJ.P.Wrners.On approximate gr ph colouring nd MAX-k -CUT algorithms based on the . -function.M nuscript,Oct. 2000.Google Scholar
- 5.L.F ybusovich.Linear systems in jordan lgebr s nd primal-dual interior point algorithms.Journal of Computational and Applied Mathematics 86:149 -175, 1997. Google ScholarDigital Library
- 6.U.Feige nd M.Goem ns.Approxim ting the v lue of two proper proof systems,with applications to MAX-2SAT nd MAX-DICUT.In Proceeding of the Third Israel Symposium on Theory of Computing and Systems pages 182 -189,1995. Google ScholarDigital Library
- 7.A.Frieze nd M.Jerrum.Improved approximation algorithms for MAX k -CUT nd MAX BISECTION. Algorithmica 18:67 -81,1997.Google Scholar
- 8.M.X.Goem ns nd D.P.Williamson.Improved approximation lgorithms for m ximum cut and s tis .bility problems using semide .nite programming.Journal of the ACM 42:1115 -1145, 1995. Google ScholarDigital Library
- 9.M.Grotschel, L.Lovasz,and A.Schrijver.The ellipsoid method nd its consequences in combin torial optimization.Combinatorica 1:169 -197,1981.Google Scholar
- 10.M.Grotschel, L.Lovasz,and A.Schrijver.Geometric Algorithms and Combinatorial Optimization Springer Verlag,Berlin,1988.Google Scholar
- 11.O.Guler.Barrier functions in interior point methods. Mathematics of Operations Research 21:860 -885, 1996. Google ScholarDigital Library
- 12.D.K rger,R.Motw ni,nd M.Sudan.Approximate graph coloring by semide .nite programming.J. ACM 45:246 -265,1998. Google ScholarDigital Library
- 13.L.Lovasz.On the Shannon capacity of graph.IEEE Transactions on Information Theory IT-25:1 -7,1979.Google ScholarCross Ref
- 14.Y.Nesterov.Qu lity of semide .nite relax tion for nonconvex quadratic optimiz tion.Technic l report, CORE,Universite Catholique de Louvain,Belgium, 1997.Google Scholar
- 15.Y.Nesterov nd A.Nemirovskii.Interior-Point Polynomial Algorithms in Convex Programming SIAM Studies in Applied Mathem tics.Society for Industri l nd Applied Mathem tics,Phil delphia, PA,1994.Google Scholar
- 16.Y.E.Nesterov and M.J.Todd.Self-scaled barriers and interior-point methods for convex programming. Mathematics of Operations Research 22:1 -42,1997. Google ScholarDigital Library
- 17.Y.E.Nesterov and M.J.Todd.Primal-dual interior-point methods for self-scaled cones.SIAM Journal on Optimization 8:324 -364,1998. Google ScholarDigital Library
- 18.S.Schmiet and F.Alizadeh.Extension of Commut tive Class of Prim l-Du l Interior Point Algorithms to Symmetric Cones.RUTCOR Rese rch Report RRR 13-99,Rutgers University,1999.Google Scholar
- 19.S.H.Schmiet nd F.Aliz deh.Associ tive lgebr s, symmetric cones nd polynomi l time interior point algorithms.RUTCOR Research Report RRR 12-99, Rutgers University,1999.Google Scholar
- 20.S.M.Selby,editor.CRC Standard Mathematical Tables The Chemical Rubber Company,Cleveland, OH,twenty-.rst edition,1973.Google Scholar
- 21.M.Skutell .Convex quadratic and semide .nite programming rel xations in scheduling.J. ACM 2000. To appear. Google ScholarDigital Library
- 22.K.-C.Toh and L.N.Trefethen.The Chebyshev polynomi ls of matrix.SIAM Journal on Matrix Analysis and Applications 20:400 -419,1998. Google ScholarDigital Library
- 23.U.Zwick.Outward rotations:A tool for rounding solutions of semide .nite programming rel xations, with applications to MAX CUT and other problems. In Proceedings of the 31st ACM Symposium on the Theory of Computing pages 679 -687,1999. Google ScholarDigital Library
Index Terms
- Approximation algorithms for MAX-3-CUT and other problems via complex semidefinite programming
Recommendations
Approximation algorithms for MAX-3-CUT and other problems via complex semidefinite programming
STOC 2001A number of recent papers on approximation algorithms have used the square roots of unity, - 1 and 1, to represent binary decision variables for problems in combinatorial optimization, and have relaxed these to unit vectors in real space using ...
Comments