Title :
Stability of a threshold-recall control self-organizing recurrent neural network
Author :
Aityan, Sergey K.
Author_Institution :
Dept. of Comput. Sci., Dortmund Univ., Germany
Abstract :
The stability of a recurrent neural network that is made up of neurons that show the activation recall dependent excitation threshold is studied. The activation function of every neuron in the network is adjusted based on the neuron activation recall. The activation threshold of the neuron is considered to be a function of the neuron activation response. Bias-adjusting neurons and activation-threshold-adjusting neurons are examined
Keywords :
recurrent neural nets; self-adjusting systems; stability; activation function; bias-adjusting neurons; neuron activation recall; stability; threshold-recall control self-organizing recurrent neural network; Artificial neural networks; Biomembranes; Computer science; Feedback control; Neural networks; Neurons; Recurrent neural networks; Stability;
Conference_Titel :
Neural Networks, 1992. IJCNN., International Joint Conference on
Conference_Location :
Baltimore, MD
Print_ISBN :
0-7803-0559-0
DOI :
10.1109/IJCNN.1992.227137
