In this lecture we will consider a particular methodology for the solution
of integer linear programming problems, known as the cutting plane method.
Cutting plane algorithms were first introduced in the `50s and have shown
their efficiency especially when applied to solving integer linear models
of combinatorial problems.
However, in this lecture we will review some procedures for the solution
of general integer linear programming problems, that is to say problems
without a particular structure.
First we are going to look at a few definitions and basic geometric properties
of integer linear programming that are at the basis of the cutting plane
approach and the general scheme of the procedure. Then we are going to
review two particular types of algorithms: the first one is based on the
Chavatal - Gomory inequalities, the second one, more recent, is based
on a convexification method and works for mixed 0-1 problems.
|
|