Konrad-zuse-zentrum F ¨ Ur Informationstechnik Berlin Improving the Accuracy of Linear Programming Solvers with Iterative Refinement Improving the Accuracy of Linear Programming Solvers with Iterative Refinement *

We describe an iterative refinement procedure for computing extended precision or exact solutions to linear programming problems (LPs). Arbitrarily precise solutions can be computed by solving a sequence of closely related LPs with limited precision arithmetic. The LPs solved share the same constraint matrix as the original problem instance and are transformed only by modification of the objective function, right-hand side, and variable bounds. Exact computation is used to compute and store the exact representation of the transformed problems, while numeric computation is used for solving LPs. At all steps of the algorithm the LP bases encountered in the transformed problems correspond directly to LP bases in the original problem description. We demonstrate that this algorithm is effective in practice for computing extended precision solutions and that this leads to direct improvement of the best known methods for solving LPs exactly over the rational numbers.

[1]  Thorsten Koch.  The final NETLIB-LP results , 2004, Oper. Res. Lett..

[2]  William J. Cook,et al.  An Exact Rational Mixed-Integer Programming Solver , 2011, IPCO.

[3]  Roland Wunderling,et al.  Paralleler und objektorientierter Simplex-Algorithmus , 1996 .

[4]  Daniel G. Espinoza.  On Linear Programming, Integer Programming and Cutting Planes , 2006 .

[5]  Dursun A. Bulutoglu,et al.  Improved WLP and GWP lower bounds based on exact integer programming , 2010 .

[6]  Michael Jünger,et al.  A branch-and-cut approach to the crossing number problem , 2008, Discret. Optim..

[7]  William J. Cook,et al.  Solving Very Sparse Rational Systems of Equations , 2011, TOMS.

[8]  Benjamin A. Burton,et al.  Computing the Crosscap Number of a Knot Using Integer Programming and Normal Surfaces , 2011, TOMS.

[9]  William J. Cook,et al.  Exact solutions to linear programming problems , 2007, Oper. Res. Lett..

[10]  Silvio Ursic,et al.  Exact Solution of Systems of Linear Equations with Iterative Methods , 1983 .

[11]  Andrea Lodi,et al.  MIPLIB 2010 , 2011, Math. Program. Comput..

[12]  Michael Kerber,et al.  ROBUST GEOMETRIC COMPUTATION , 2012 .

[13]  William J. Cook,et al.  Safe Lower Bounds for Graph Coloring , 2011, IPCO.

[14]  Fernando Ordóñez,et al.  Computational Experience and the Explanatory Value of Condition Measures for Linear Optimization , 2003, SIAM J. Optim..

[15]  Zhendong Wan,et al.  An algorithm to solve integer linear systems exactly using numerical methods , 2006, J. Symb. Comput..

[16]  Kurt Mehlhorn,et al.  Certifying and repairing solutions to large LPs how good are LP-solvers? , 2003, SODA '03.

[17]  F. M. D. O. Filho,et al.  Fourier analysis, linear programming, and densities of distance avoiding sets in R^n , 2008, 0808.1822.

[18]  Robert H. Halstead,et al.  Matrix Computations , 2011, Encyclopedia of Parallel Computing.

[19]  Yuval Rabani,et al.  Linear Programming , 2007, Handbook of Approximation Algorithms and Metaheuristics.

[20]  Xiaomei Yang.  Rounding Errors in Algebraic Processes , 1964, Nature.

[21]  T. Hales.  The Kepler conjecture , 1998, math/9811078.

[22]  B. David Saunders,et al.  Numeric-symbolic exact rational linear system solver , 2011, ISSAC '11.

[23]  Illya V. Hicks,et al.  The branchwidth of graphs and their cycle matroids , 2007, J. Comb. Theory, Ser. B.

[24]  Chee-Keng Yap,et al.  Robust Geometric Computation , 2016, Encyclopedia of Algorithms.

[25]  Victor Y. Pan,et al.  Nearly optimal solution of rational linear systems of equations with symbolic lifting and numerical initialization , 2011, Comput. Math. Appl..