Title :
Fault tolerance parameter model of radial basis function networks
Author :
Mallofré, áAndreu Catal ; Llanas, Xavier Parra
Author_Institution :
Dept. d´´Enginyeria de Sistemes, Autom. i Inf. Ind., Univ. Politecnica de Catalunya, Barcelona, Spain
Abstract :
An ANN fault tolerance study needs a realistic work of the faults influence on the network´s performance. A parametric fault model and multiple faults seem to be closer to possible VLSI implementation problems. A methodology of a worst case fault model was applied to radial basis function networks showing up to ±20% of process tolerance parameter
Keywords :
feedforward neural nets; fault tolerance parameter model; multiple faults; parametric fault model; radial basis function networks; worst case fault model; Artificial neural networks; Biological system modeling; Degradation; Electronic mail; Fault tolerance; Fault tolerant systems; Radial basis function networks; Redundancy; System performance; Very large scale integration;
Conference_Titel :
Neural Networks, 1996., IEEE International Conference on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-3210-5
DOI :
10.1109/ICNN.1996.549101
