Modelling with Integer Variables.
Abstract
Representing nonlinear optimization problems as mixed-integer programs has largely been considered as; (1) an art with few unresolved theoretical issues, and (2) a fairly standard preprocessing routine when combined with several ad hoc modelling improvements which have evolved through computational experience. The more common avenue of research in mixed-integer programming has focused upon finding improved algorithms and heuristics to solve the problems, assuming a standard mixed-integer representation exists. This thesis, takes a step backwards and re-examines the theoretical issues and subtleties involved in representing problems with mixed-integer representations.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1984
- Accession Number
- ADA140212
Entities
People
- J. K. Lowe
Organizations
- Air Force Institute of Technology