Title :
Mutual information in a linear noisy network
Author :
Campa, Alessandro ; Giudice, Paolo Del ; Parga, Nestor ; Nadal, J.-P.
Author_Institution :
Lab. of Phys., Istituto Nazionale di Fisica Nucl., Rome, Italy
fDate :
31 Aug-2 Sep 1995
Abstract :
Considers a linear, one-layer feedforward neural network performing a coding task under noisy conditions. The authors determine the family of synaptic couplings that maximizes the mutual information between input and output distribution. Optimisation is performed under different constraints on the synaptic efficacies. The authors analyze the dependence of the solutions on input and output noises
Keywords :
encoding; feedforward neural nets; noise; optimisation; probability; self-adjusting systems; coding task; input noises; linear noisy network; linear one-layer feedforward neural network; mutual information; output noises; synaptic couplings; synaptic efficacies; Feedforward neural networks; Information theory; Intelligent networks; Laboratories; Mutual coupling; Mutual information; Neural networks; Neurons; Physics; Probability distribution;
Conference_Titel :
Neural Networks for Signal Processing [1995] V. Proceedings of the 1995 IEEE Workshop
Conference_Location :
Cambridge, MA
Print_ISBN :
0-7803-2739-X
DOI :
10.1109/NNSP.1995.514892
