cplexqp, when used for nonconvex quadratic programming, is giving the same results (fvalue) when tried with different random initial points. Actually, different initial points should make the algorithm converge to different local minima, I believe? Why doesnt this happen?
Thanks in advance
Pinned topic Starting point for nonconvex quadratic programming
Answered question This question has been answered.
Unanswered question This question has not been answered yet.
Updated on 2011-11-22T14:46:13Z at 2011-11-22T14:46:13Z by SystemAdmin