Abstract
We present branching-on-hyperplane methods for solving mixed integer linear and mixed integer convex programs. In particular, we formulate the problem of finding a good branching hyperplane using a novel concept of adjoint lattice. We also reformulate the problem of rounding a continuous solution to a mixed integer solution. A worst case complexity of a Lenstra-type algorithm is established using an approximate log-barrier center to obtain an ellipsoidal rounding of the feasible set. The results for the mixed integer convex programming also establish a complexity result for the mixed integer second order cone programming and mixed integer semidefinite programming feasibility problems as a special case. Our results motivate an alternative reformulation technique and a branching heuristic using a generalized (e.g., ellipsoidal) norm reduced basis available at the root node.
Original language | English (US) |
---|---|
Pages (from-to) | 623-649 |
Number of pages | 27 |
Journal | Journal of Global Optimization |
Volume | 49 |
Issue number | 4 |
DOIs | |
State | Published - Apr 2011 |
Keywords
- Analytic center
- Convex programming
- Interior point methods
- Lattice basis reduction
- Linear programming
- Mixed integer programming
- Volumetric center
ASJC Scopus subject areas
- Computer Science Applications
- Management Science and Operations Research
- Control and Optimization
- Applied Mathematics