Title :
Sub-gradient based projection neural networks for non-differentiable optimization problems
Author :
Li, Guo-cheng ; Dong, Zhi-Ling
Author_Institution :
Dept. of Math., Beijing Inf. Sci. & Technol. Univ., Beijing
Abstract :
This paper further investigates the sub-gradient projection neural networks model for solving non- differentiable convex optimization problems proposed by Li et al. (2006). It is proved in this paper that when the initial points are belong to the constraint set or the initial points are not belong to the constraint set and the objective function is strictly convex, the network trajectories converge to an optimal solution of the primal optimal problem.
Keywords :
gradient methods; neural nets; optimisation; network trajectories; nondifferentiable optimization problems; objective function; primal optimal problem; sub-gradient based projection neural networks; Constraint optimization; Cybernetics; Hopfield neural networks; Information science; Linear programming; Machine learning; Mathematical model; Mathematics; Neural networks; Virtual colonoscopy; Differential inclusions; Projection neural network; Sub-gradient;
Conference_Titel :
Machine Learning and Cybernetics, 2008 International Conference on
Conference_Location :
Kunming
Print_ISBN :
978-1-4244-2095-7
Electronic_ISBN :
978-1-4244-2096-4
DOI :
10.1109/ICMLC.2008.4620520
