Muutke küpsiste eelistusi

Exploring Interior-point Linear Programming: Algorithms and Software [Pehme köide]

  • Formaat: Paperback / softback, 235 pages, kõrgus x laius x paksus: 229x175x12 mm, kaal: 431 g, 39, Contains 1 Paperback / softback and 1 Diskette
  • Sari: Foundations of Computing
  • Ilmumisaeg: 10-Nov-1993
  • Kirjastus: MIT Press
  • ISBN-10: 0262510731
  • ISBN-13: 9780262510738
  • Formaat: Paperback / softback, 235 pages, kõrgus x laius x paksus: 229x175x12 mm, kaal: 431 g, 39, Contains 1 Paperback / softback and 1 Diskette
  • Sari: Foundations of Computing
  • Ilmumisaeg: 10-Nov-1993
  • Kirjastus: MIT Press
  • ISBN-10: 0262510731
  • ISBN-13: 9780262510738
Commonly known as Karmarkar's algorithm, the interior-point technique is proving especially powerful for the solution of large-scale linear programming problems, with better performance bounds than the simplex algorithm. Delta Airlines, for example, uses interior-point methods to schedule its air crews and planes. A software package and user's guide is included with IBM PC implementations of the Primal, Dual, and Primal-Dual interior-point linear programming algorithms. Annotation copyright Book News, Inc. Portland, Or.

Linear programming is widely used in industry to solve complex planning and resourceallocation problems. This book provides practitioners as well as students of this generalmethodology with an easily accessible introduction to the new class of algorithms known asinterior-point methods for linear programming. In addition to presenting the theoretical andalgorithmic background necessary for dealing with specific interior-point linear programmingalgorithms, it offers a review of modeling linear programming problems, a review of the simplexalgorithm that has been used to solve linear programming problems in the past, and a complete user'sguide to the software that is included with the book.The interior-point technique is provingespecially powerful for the solution of large-scale linear programming problems, with betterperformance bounds than the simplex algorithm. For example, the U.S. Military airlift command hassolved their scheduling problem using interior-point algorithms much faster and with a longerplanning horizon than was possible with the simplex algorithms, and Delta expects to save millionsof dollars by using interior-point methods to schedule their air crews and planes.The softwarepackage is designed for use on IBM-PC microcomputers (and compatibles), a platform that provides anideal environment for students of linear programming interested in exploring and studying these newalgorithms.Contents: Preparations. Introduction. Modeling Linear Optimization Problems. The SimplexAlgorithm. A First Look at an Interior Point Algorithm. Algorithms. The Primal Algorithm. The DualAlgorithm. The Primal-Dual Algorithm. Implementation Issues. Solutions. The Integrated Environment.Command Line Operations. Appendixes.



This book provides practitioners as well as students of this general methodology withan easily accessible introduction to the new class of algorithms known as interior-point methods forlinear programming.

Series Foreword xi
Preface xiii
List of Notations
xvii
Up and Running xix
I Preparations
Introduction
3(4)
Modeling Linear Optimization Problems
7(14)
Linear Programs in Standard Form
8(4)
Modeling Issues and Nonlinear Effects
12(2)
Graphical Solution of Linear Programs
14(3)
Problem Areas
17(3)
Additional Reading
20(1)
The Simplex Algorithm
21(32)
Linear Systems of Equations
22(1)
Bases and Basic Solutions
23(2)
Replacing Vectors in a Basis
25(2)
Fundamental Properties of Linear Programming Problems
27(8)
Pivot Operation in Solving a System of Equations
35(7)
Starting and Stopping the Simplex Algorithm
42(5)
Matrix Form of the Pivot Operation
47(2)
Duality
49(3)
Additional Reading
52(1)
A First Look at an Interior-Point Algorithm
53(20)
Geometric Explorations
53(7)
Mathematical Requirements
60(9)
Additional Reading
69(4)
II Algorithms
The Primal Algorithm
73(18)
Scaling
74(3)
Taking an Interior Step
77(1)
Free Variables
78(1)
How to Start
79(2)
When to Stop
81(2)
A Digression
83(6)
Additional Reading
89(2)
The Dual Algorithm
91(12)
Scaling
92(3)
Primal Estimates
95(6)
Additional Reading
101(2)
The Primal-Dual Algorithm
103(22)
Newton's Method
105(2)
The Barrier Method
107(1)
Developing the Algorithm
108(12)
Summary
120(4)
Additional Reading
124(1)
Implementation Issues
125(28)
Solving Systems of Linear Equations
125(5)
Solving Symmetric Systems of Equations
130(13)
Sparse Matrix Techniques
143(7)
Additional Reading
150(3)
III Solutions
The Integrated Environment
153(32)
The File Command
155(3)
The Edit Command
158(5)
The Spec File Command
163(11)
The Solve Command
174(1)
The View Command
175(2)
The Print Command
177(3)
The Run Demo Command
180(5)
Command-Line Operations
185(4)
Solving a Model
185(1)
Editing the Specfile
185(4)
Appendix A: The MPS File Format 189(6)
Appendix B: The Netlib Test Collection 195(8)
Selected Bibliography 203(4)
Index 207