Patent
1995-07-19
1997-06-17
Davis, George B.
395 22, 395 23, G06F 1518
Patent
active
056404948
ABSTRACT:
A neural network comprises an input port connected to an output port by one or more paths, each of which comprises an alternating series of weights and neurons. The weights amplify passing signals by a strength factor. The network can be trained by finding a set of strength factor values for the weights such that the network produces the correct output pattern from a given input pattern. During training, a strength factor perturbating and refresh means applies perturbations to the strength factors of weights in the network, and updates the values of the strength factors depending on the difference between signals appearing at the output port, for a given pair of input and training patterns, when the weight is perturbed, and when it is not.
REFERENCES:
patent: 4885757 (1989-12-01), Provence
patent: 4912647 (1990-03-01), Wood
patent: 4958939 (1990-09-01), Samad
patent: 5050095 (1991-09-01), Samad
patent: 5075868 (1991-12-01), Andes
patent: 5083285 (1992-01-01), Shima
patent: 5103496 (1992-04-01), Andes et al.
patent: 5150450 (1992-09-01), Swenson et al.
patent: 5157738 (1992-10-01), Carpenter et al.
patent: 5175797 (1992-12-01), Funabashi et al.
patent: 5255346 (1993-10-01), Wu et al.
patent: 5293453 (1994-03-01), Frazier
Dembo et al, "Model-Free Distributed Learning", IEEE Trans. on Neural Networks. vol. 1. No. 1. Mar. 1990.
Flower Barry Glen
Jabri Marwan Anwar
Davis George B.
The University of Sydney
LandOfFree
Neural network with training by perturbation does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Neural network with training by perturbation, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Neural network with training by perturbation will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2164413