## An Interior Point Approach to the Maximum Independent Set Problem in Dense Random Graphs (1989)

Venue: | In Proceedings of the XIII Latin American Conference on Informatics |

Citations: | 6 - 3 self |

### BibTeX

@INPROCEEDINGS{Karmarkar89aninterior,

author = {Narendra Karmarkar and Mauricio G.C. Resende and K. G. Ramakrishnan},

title = {An Interior Point Approach to the Maximum Independent Set Problem in Dense Random Graphs},

booktitle = {In Proceedings of the XIII Latin American Conference on Informatics},

year = {1989},

pages = {241--260}

}

### OpenURL

### Abstract

We present an interior point approach to the zero-one integer programming feasibility problem based on the minimization of an appropriate potential function. Given a polytope defined by a set of linear inequalities, this procedure generates a sequence of strict interior points of this polytope, such that each consecutive point reduces the value of the potential function. An integer solution (not necessarily feasible) is generated at each iteration by a rounding scheme. The direction used to determine the new iterate is computed by solving a nonconvex quadratic program on an ellipsoid. We illustrate the approach by considering a class of difficult NP-complete problems: finding a maximum independent set of a dense random graph. Some implementation details are discussed and preliminary computational results are presented. We solve several large independent set problems in graphs having up to 1000 vertices and over 250,000 edges. Key words: Integer programming, interior point method, maxim...

### Citations

11502 |
Computers and Intractability, A Guide to the Theory of NPCompleteness
- Garey, Johnson
- 1979
(Show Context)
Citation Context ...of �� G if and only if V n S is a minimum vertex cover of �� G. Finding a maximum independent set, a maximum clique or a minimum vertex cover of a graph are equivalent and are known to be NP-c=-=omplete [7]-=-. Practical applications of these models are abundant, e.g. information retrieval, signal transmission analysis, classification theory, economics, scheduling, experimental design and computer vision (... |

1934 | Random Graphs
- Bollobás
- 1985
(Show Context)
Citation Context ...lts are preliminary, since further implementation is ongoing. We have selected this class of problems because it is easy to generate hard instances using some results from the theory of random graphs =-=[4]. Consider a-=- graph G = (V; E) with vertex set V and edge set E and its complement, �� G = (V; �� E), where �� E = f(i; j) 62 E; i; j 2 V; i 6= jg. An independent set of vertices (or vertex packing or ... |

1014 |
An algorithm for least-squares estimation of nonlinear parameters
- Marquardt
- 1963
(Show Context)
Citation Context ...ial time algorithm for nonconvex quadratic programming on an ellipsoid that differs from the approach to be presented here. Our approach is similar to the classical Levenberg-Marquardt methods, [11], =-=[12]-=-, first suggested in the context of nonlinear least squares. The following proposition describes a suitable inscribed ellipsoid. Proposition 2.1 Consider the polytope defined as L = n w 2 ! m jA T wsc... |

221 |
The zero error capacity of a noisy channel
- Shannon
- 1956
(Show Context)
Citation Context ...hese models are abundant, e.g. information retrieval, signal transmission analysis, classification theory, economics, scheduling, experimental design and computer vision (See [1], [2], [3], [5], [6], =-=[15]-=-, [9] and [16] for details). A simple integer programming formulation for vertex packing is given next. Let S ae V be an independent k-set of the graph G = (V; E) and define w j = 8 ! : 1 if vertex j ... |

177 |
A method for the solution of certain problems in least squares
- Levenberg
- 1944
(Show Context)
Citation Context ...olynomial time algorithm for nonconvex quadratic programming on an ellipsoid that differs from the approach to be presented here. Our approach is similar to the classical Levenberg-Marquardt methods, =-=[11]-=-, [12], first suggested in the context of nonlinear least squares. The following proposition describes a suitable inscribed ellipsoid. Proposition 2.1 Consider the polytope defined as L = n w 2 ! m jA... |

141 |
Graph theory with applications to engineering and computer science
- Deo
- 1974
(Show Context)
Citation Context ... of these models are abundant, e.g. information retrieval, signal transmission analysis, classification theory, economics, scheduling, experimental design and computer vision (See [1], [2], [3], [5], =-=[6]-=-, [15], [9] and [16] for details). A simple integer programming formulation for vertex packing is given next. Let S ae V be an independent k-set of the graph G = (V; E) and define w j = 8 ! : 1 if ver... |

71 |
Theory of Graphs and its Applications
- Berge
- 1958
(Show Context)
Citation Context ...plications of these models are abundant, e.g. information retrieval, signal transmission analysis, classification theory, economics, scheduling, experimental design and computer vision (See [1], [2], =-=[3]-=-, [5], [6], [15], [9] and [16] for details). A simple integer programming formulation for vertex packing is given next. Let S ae V be an independent k-set of the graph G = (V; E) and define w j = 8 ! ... |

66 |
Computationally related problems
- Sahni
- 1974
(Show Context)
Citation Context ...One way of selecting (-1,1) integer solutions over fractional solutions in linear programming is to introduce the quadratic objective function maximize w T w = m X i=1 w 2 i and solve the NP-complete =-=[14]-=- nonconvex quadratic programming problem (QUADRATIC PROGRAMMING) maximize w T w (3) subject to : A T wsc (4) Note that an upper bound on the objective function (3) is m. The following proposition esta... |

18 |
An interior point approach to NP-complete problems
- Karmarkar
(Show Context)
Citation Context ...ation. Finally we illustrate how this algorithm can be applied to a difficult combinatorial problem: Finding an independent k-set of a dense random graph. 2. Nonconvex Quadratic Programming Karmarkar =-=[10]-=- suggested an interior point approach to NP-complete problems. In this section we provide an algorithm in pseudo-code form detailing that approach. We also provide details on applying this algorithm t... |

13 |
Probabilistic bounds and heuristic algorithms for coloring large random graphs
- Johri, Matula
- 1982
(Show Context)
Citation Context ...ph, the resulting graph is very small. By enumeration, a maximum independent set of the reduced graph is readily available. For enumeration we use the semi-exhaustive greedy independent set algorithm =-=[8]-=- implemented by D.S. Johnson with parameter settings to do exhaustive enumeration. At each call of procedure descent direction a linear system M \Deltaw = b must be solved at least once (lines 4-8). W... |

6 |
Economic applications of the theory of graphs
- Avondo-Bodeno
- 1962
(Show Context)
Citation Context ...actical applications of these models are abundant, e.g. information retrieval, signal transmission analysis, classification theory, economics, scheduling, experimental design and computer vision (See =-=[1]-=-, [2], [3], [5], [6], [15], [9] and [16] for details). A simple integer programming formulation for vertex packing is given next. Let S ae V be an independent k-set of the graph G = (V; E) and define ... |

4 |
and Chang Sung Yu. Finding a maximum clique in an arbitrary graph
- Balas
- 1986
(Show Context)
Citation Context ...al applications of these models are abundant, e.g. information retrieval, signal transmission analysis, classification theory, economics, scheduling, experimental design and computer vision (See [1], =-=[2]-=-, [3], [5], [6], [15], [9] and [16] for details). A simple integer programming formulation for vertex packing is given next. Let S ae V be an independent k-set of the graph G = (V; E) and define w j =... |

4 |
On the interior algorithms for nonconvex quadratic programming
- Ye
- 1988
(Show Context)
Citation Context ...amma 1 f 0 w + 1 n AD \Gamma1 e: (7) Minimizing (5) subject to A T wsc is NP-complete. However, if the polytope is substituted by an inscribed ellipsoid, the resulting approximate problem is easy. Ye =-=[17]-=- has proposed a polynomial time algorithm for nonconvex quadratic programming on an ellipsoid that differs from the approach to be presented here. Our approach is similar to the classical Levenberg-Ma... |

3 |
Solution characteristics and algorithms for the vertex packing problem
- Jr
- 1973
(Show Context)
Citation Context ...odels are abundant, e.g. information retrieval, signal transmission analysis, classification theory, economics, scheduling, experimental design and computer vision (See [1], [2], [3], [5], [6], [15], =-=[9]-=- and [16] for details). A simple integer programming formulation for vertex packing is given next. Let S ae V be an independent k-set of the graph G = (V; E) and define w j = 8 ! : 1 if vertex j 2 S \... |

3 |
A survey of progress in graph theory in the Soviet Union
- Turner, Kautz
- 1970
(Show Context)
Citation Context ...e abundant, e.g. information retrieval, signal transmission analysis, classification theory, economics, scheduling, experimental design and computer vision (See [1], [2], [3], [5], [6], [15], [9] and =-=[16]-=- for details). A simple integer programming formulation for vertex packing is given next. Let S ae V be an independent k-set of the graph G = (V; E) and define w j = 8 ! : 1 if vertex j 2 S \Gamma1 ot... |

2 | A global router based on Karmarkar's interior point method - Pai, Karmarkar, et al. - 1988 |

1 |
De l'utilisation de la notion de clique en mati`ere de typologie des populations
- Degot, Hualde
- 1975
(Show Context)
Citation Context ...tions of these models are abundant, e.g. information retrieval, signal transmission analysis, classification theory, economics, scheduling, experimental design and computer vision (See [1], [2], [3], =-=[5]-=-, [6], [15], [9] and [16] for details). A simple integer programming formulation for vertex packing is given next. Let S ae V be an independent k-set of the graph G = (V; E) and define w j = 8 ! : 1 i... |

1 |
Economic applications of the theory of graphs. Gordon&Breach Science
- Avondo-Bodeno
- 1962
(Show Context)
Citation Context ...actical applications of these models are abundant, e.g. information retrieval, signal transmission analysis, classification theory, economics, scheduling, experimental design and computer vision (See =-=[1]-=-, [2], [3], [5], [6], [15], [9] and [16] for details). A simple integer programming formulation for vertex packing is given next. Let S ⊂ V be an independent k-set of the graph G =(V,E) and define ⎧ ⎨... |