Branch and cut methods for network optimization
Mathematical and Computer Modelling: An International Journal
| Hi-index | 0.00 |
A cutting plane algorithm is presented for solving combinatorial linear programs---integer programs with 0/1 vertices represented by a separation oracle. The algorithm is a standard cutting plane method but uses a prescribed dual criterion for choosing a cut at each iteration. As a result, it is possible to demonstrate that for problems containing a ball whose size is polynomially bounded from below, there exists an algorithm polynomial in the running time of the separation oracle and pseudopolynomial in the size of the objective function. In particular, the cardinality versions of many combinatorial optimization problems are shown to be solvable in polynomial time using this generic algorithm.