Recurrent Project Neural Network for solving Semidefinite Programming

Leifu Gao, Changli Feng, Haiyan Wei, Xiaokai Chang
2010 2010 3rd International Conference on Computer Science and Information Technology  
In order to solve the Semidefinite Programming, this paper introduces the recurrent neural network model. By introducing the equivalent variational inequality problem and the neural network which has the same solution with the variational inequality, the model converges to the solution of semidefinite programming. At the same time, numerical simulation also proves it very well. In short, neural network is an effective way to solve semidefinite programming, so as to provide a viable option for engineering.
doi:10.1109/iccsit.2010.5563770 fatcat:o6t4e5colnhf3atezrz4krk2gm