Supervised training algorithms for B-Spline neural networks and neuro-fuzzy systems

Ruano, Antonio; Cabrita, Cristiano Lourenço; Oliveira, J. V.; Kóczy, László T.

http://hdl.handle.net/10400.1/2338

Use this identifier to reference this record.

Name:	Description:	Size:	Format:
ruano 2002.pdf		1.07 MB	Adobe PDF	Download

Send Feedback

Authors

Ruano, Antonio

Cabrita, Cristiano Lourenço

Oliveira, J. V.

Kóczy, László T.

Abstract(s)

Complete supervised training algorithms for B-Spline neural networks and fuzzy rulebased systems are discussed. By introducing the relationships between B-Spline neural networks and Mamdani (satisfying certain assumptions) and Takagi±Kang±Sugeno fuzzy models, training algorithms developed initially for neural networks can be adapted to fuzzy systems. The standard training criterion is reformulated, by separating its linear and nonlinear parameters. By employing this reformulated criterion with the Levenberg-Marquardt algorithm, a new training method, offering a fast rate of convergence is obtained. It is also shown that the standard Error-Back Propagation algorithm, the most common training method for this class of systems, exhibits a very poor and unreliable performance.

URI

http://hdl.handle.net/10400.1/2338

Citation

Ruano, A. E.; Cabrita, C.; Oliveira, J. V.; Kóczy, L. T. Supervised training algorithms for B-Spline neural networks and neuro-fuzzy systems, International Journal of Systems Science, 33, 8, 689-711, 2002.