Radial basis function networks with quantized parameters

• Autor(a) principal: Lucks, Marcio B. [UNESP]
• Data de Publicação: 2008
• Outros Autores: Nobuo, Oki [UNESP]
• Tipo de documento: Artigo de conferência
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1109/CIMSA.2008.4595826; http://hdl.handle.net/11449/70591
• topic: Function approximation; Quantized parameters; Radial basis function network; Artificial intelligence; Chlorine compounds; Feedforward neural networks; Intelligent control; Networks (circuits); Polynomial approximation; Approximation properties; Circuit complexity; Computational intelligence; International conferences; Low-power applications; Measurement systems; Memory size; Mixed-signal circuits; Network parameters; Quantization levels; Simulation results; Sinusoidal functions; Radial basis function networks

A RBFN implemented with quantized parameters is proposed and the relative or limited approximation property is presented. Simulation results for sinusoidal function approximation with various quantization levels are shown. The results indicate that the network presents good approximation capability even with severe quantization. The parameter quantization decreases the memory size and circuit complexity required to store the network parameters leading to compact mixed-signal circuits proper for low-power applications. ©2008 IEEE.