EFFICIENT SUBOPTIMAL ALGORITHMS FOR INTEGER LINEAR PROGRAMMING WITH AN INTERIOR.
STANFORD UNIV CALIF DEPT OF INDUSTRIAL ENGINEERING
Pagination or Media Count:
The paper presents and evaluates some new suboptimal algorithms for the pure integer linear programming problem having only inequality constraints. The computation time required by these algorithms after obtaining the optimal noninteger solution has been only a small fraction of that required by the simplex method. Furthermore, the solution obtained by the better algorithms consistently has been close to optimal and frequently has actually been optimal. Plans for generalizing these algorithms also are outlines. A companion paper presents an optimal bound-and-scan algorithm that would be used in conjuection with these suboptimal algorithms. Author
- Operations Research