Tuesday, 13:15 - 13:40 h, Room: H 2053

 

Hermann Schichl
Balanced rigorous relaxation methods in global optimization

Coauthors: Mihaly Cs Markot, Arnold Neumaier

 

Abstract:
Relaxations are an important tool for solving global optimization problems. Linear and more general convex relaxations are most commonly employed in global optimization algorithms. We present a new relaxation scheme for mixed integer nonlinear programs which balances the dimension of the relaxed problem with its enclosure properties and apply it to generate LP and MIP relaxations of a factorable global optimization problem as well as convex QP and QCQP relaxations.
We generate these relaxations by analyzing the structure of the operational directed acyclic graph of the problem and use a combination of local relaxation at a node of the graph and slope based relaxation methods working on subgraphs. This allows to limit the size of the relaxation, hereby reducing the computational effort for the solution of the relaxations during the branch and bound process.

 

Talk 1 of the invited session Tue.2.H 2053
"Rigorous global optimization" [...]
Cluster 9
"Global optimization" [...]

 

  online cash advance . If you have already decided to take Levitra, be sure to consult a doctor, you don't have any contraindications and act strictly due to a prescription.