Title :
A VLSI friendly neural network with localised transfer functions
Author :
Körner, Tim ; Rückert, Ulrich ; Geva, Shlomo ; Malmstrom, Kurt ; Sitte, Joaquin
Author_Institution :
Dept. of Tech. Electron., Tech. Univ. Hamburg-Harburg, Germany
Abstract :
The local cluster (LC) artificial neural net architecture performs as well as the radial basis functions networks as a computational method for multidimensional function approximation. The LC network combines sigmoidal neurones in clusters that have localised response in input space. This construction gives the LC nets additional flexibility over RBF nets and makes them more VLSl friendly. We investigate the computational abilities of two versions of the LC architecture and confirm the feasibility of an analog implementation by showing a circuit design verified by SPICE simulation
Keywords :
VLSI; analogue processing circuits; circuit CAD; function approximation; neural chips; neural net architecture; transfer functions; SPICE simulation; analog VLSI chip; digital storage; local cluster neural net; localised transfer functions; multidimensional function approximation; neural net architecture; sigmoidal neurones; Analog computers; Artificial neural networks; Computer architecture; Computer networks; Function approximation; Multidimensional systems; Neural networks; Radial basis function networks; Transfer functions; Very large scale integration;
Conference_Titel :
Neural Networks, 1995. Proceedings., IEEE International Conference on
Conference_Location :
Perth, WA
Print_ISBN :
0-7803-2768-3
DOI :
10.1109/ICNN.1995.488087
