A target oriented averaging search trajectory and its application in artificial neural networks

The training of artificial neural networks usually involves nonsmooth objective func- tions to be minimized. This optimization problem is currently solved just avoing saddle points and thus reaching a local and fortunately a global minimizer. Well-known opti- mizers like Gradient Descent (GD) or its...

• Hoofdauteur: Rojas Jiménez, Ángel Adrián (author)
• Formaat: bachelorThesis
• Taal: eng
• Gepubliceerd in: 2019
• Onderwerpen: Successive Abs-Linearization; Quadratic Regularization; AbsNormal Form; Generalized descent; Dynamic Search Trajectory; Abs-Linearización Sucesiva; Regularización Cuadrática; Forma Abs-Normal; Descenso Generalizado; Trayectoria de búsqueda dinámica
• Online toegang: http://repositorio.yachaytech.edu.ec/handle/123456789/104