Absolute stability (ABST) was applied to special classes of nonsymmetric Hopfield neural networks and other neural variations. For ABST to exist in a neural network, there must be a unique equilibrium point attracting all trajectories of motion. This equilibrium must be valid for all neuron activation functions and all constant neural network inputs. Results of the application revealed that ABST was applicable to special classes of nonsymmetric Hopfield neural networks. Necessary and sufficient conditions for ABST were also found for Hopfield-type networks & composite neural networks. The two networks contained variable & constraint neurons. Such type of neurons are useful in solving linear and quadratic programming problems.
Forti, M., Liberatore, A., Manetti, S., Marini, M. (1994). On absolute stability of neural networksProceedings of IEEE International Symposium on Circuits and Systems - ISCAS '94. In Proceedings of IEEE International Symposium on Circuits and Systems - ISCAS '94 (pp.241-244). IEEE [10.1109/ISCAS.1994.409572].
On absolute stability of neural networksProceedings of IEEE International Symposium on Circuits and Systems - ISCAS '94
Forti, M.;
1994-01-01
Abstract
Absolute stability (ABST) was applied to special classes of nonsymmetric Hopfield neural networks and other neural variations. For ABST to exist in a neural network, there must be a unique equilibrium point attracting all trajectories of motion. This equilibrium must be valid for all neuron activation functions and all constant neural network inputs. Results of the application revealed that ABST was applicable to special classes of nonsymmetric Hopfield neural networks. Necessary and sufficient conditions for ABST were also found for Hopfield-type networks & composite neural networks. The two networks contained variable & constraint neurons. Such type of neurons are useful in solving linear and quadratic programming problems.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/11365/30273
Attenzione
Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo
