Title :
Control of generalization with a bi-objective sliding mode control algorithm
Author :
Costa, Marcelo Azevedo ; Braga, Antônio Pádua ; De Menezes, Benjamin Rodrigues ; Parma, Gustavo Guimarães ; Teixeria, Rde.A.
Author_Institution :
Univ. Fed. de Minas Gerais, Belo Horizonte, Brazil
Abstract :
This paper presents a new sliding mode control algorithm that is able to guide the trajectory of a multilayer perceptron within the plane formed by the two objectives: training set error and norm of the weight vectors. The results show that the neural networks obtained are able to generate the Pareto set, from which a model with the smallest validation error is selected.
Keywords :
generalisation (artificial intelligence); learning (artificial intelligence); multilayer perceptrons; optimisation; variable structure systems; Pareto set; generalization; multilayer perceptron; neural networks; sliding mode control; training set error; weight vectors; Backpropagation; Character generation; Constraint optimization; Error correction; Minimization methods; Multilayer perceptrons; Neural networks; Pareto optimization; Sliding mode control; State-space methods;
Conference_Titel :
Neural Networks, 2002. SBRN 2002. Proceedings. VII Brazilian Symposium on
Print_ISBN :
0-7695-1709-9
DOI :
10.1109/SBRN.2002.1181432
