A Further Investigation of Efficient Heuristic Procedures for Integer Linear Programming with an Interior.
Abstract
This paper presents the results of an extensive investigation of algorithmic heuristic procedures for general pure integer linear programming problems having only inequality constraints. Included are a number of promising new variations and extensions of the procedures previously proposed by the author. Extensive computational experimentation has largely succeeded in identifying a flexible package of the most effective approaches, ranging from a very fast streamlined procedure to a very powerful combination of procedures. These procedures are both extremely efficient (comparable to the simplex method) and very effective in identifying good solutions (often obtaining an optimal one). Although they are designed primarily for dealing algorithmically with the frequently encountered problems that are too large to be computationally feasible for exact algorithms, they also can be valuable on smaller problems by quickly providing an advanced starting solution for such algorithms. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1977
- Accession Number
- ADA049593
Entities
People
- Frederick Stanton Hillier
Organizations
- Stanford University