Efficient DC programming approaches for the asymmetric eigenvalue complementarity problem
Y. S. Niu, T. Pham Dinh, H. A. Le Thi and J. J. Júdice
Abstract
In
this paper, we propose nonlinear programming formulations (NLP) and DC
(Difference of Convex functions) programming approaches for the
asymmetric eigenvalue complementarity problem (EiCP). The EiCP has a
solution if and only if these NLPs have zero global optimal value. We
reformulate the NLPs as DC Programs (DCP) which can be efficiently
solved by DCA (DC Algorithm). Some preliminary numerical results
illustrate the good performance of the proposed methods.