Abstract |
The quadratic assignment problem (QAP), as one of fundamental combinatorial optimization problems, has wide applications in graph matching, computer manufacturing, scheduling, etc. In this talk, we shall consider the Lp regularization for QAP with 0<p<1 or p>1. Some theoretic properties of our Lp regularization will be introduced and analyzed. The preliminary numerical results on QAPLIB instances show that our Lp regularization approach is very promising for QAP. |