DocumentCode
274146
Title
Probabilistic learning on a neural network
Author
Wright, W.A.
Author_Institution
Sowerby Res. Centre, British Aerosp., UK
fYear
1989
fDate
16-18 Oct 1989
Firstpage
153
Lastpage
157
Abstract
This paper presents an information theoretic description of a possible learning mechanism for an adaptive signal processing system. It is shown that this learning system, when applied to a multilayered perceptron, produces a learning algorithm which closely resembles the conventional generalised delta rule, but treats the outputs of the network wholly as probabilities
Keywords
adaptive systems; information theory; learning systems; neural nets; probability; signal processing; adaptive signal processing system; generalised delta rule; information theory; multilayered perceptron; neural network; probabilistic learning;
fLanguage
English
Publisher
iet
Conference_Titel
Artificial Neural Networks, 1989., First IEE International Conference on (Conf. Publ. No. 313)
Conference_Location
London
Type
conf
Filename
51950
Link To Document