THE HOLOGRAPHIC NEURAL NETWORK PERFORMANCE COMPARISON WITH OTHER NEURAL NETWORKS

PDF

Authors
  1. Klepko, R.
Corporate Authors
Defence Research Establishment Ottawa, Ottawa ONT (CAN)
Abstract
The Airborne Radar Section at the Defence Research Establishment Ottawa is interested in determining and evaluating techniques which may be used to recognize high resolution radar images of ships. Artificial Neural Networks attempt to mimic the operation of the human brain to perform tasks such as speech understanding, robot control and pattern recognition. These networks do not make any assumptions about the distribution statistics of the input data that is processed. Some networks must be trained, while others can learn on their own. These networks can learn data which is uncorrupted, but recall the same data after it has been buried in noise. The Holographic Neural Network (HNN), promises a very large data storage capacity and excellent generalization capability. Both of these attributes can be achieved with only a few learning trials, which is unlike most neural networks that require on the order of thousands of learning trials. TRUNCATED
Report Number
DREO-TN-91-18 — Technical Note
Date of publication
15 Oct 1991
Number of Pages
37
DSTKIM No
92-00969
CANDIS No
103862
Format(s):
Hardcopy;Originator's fiche received by DSIS;Document Image stored on Optical Disk

Permanent link

Document 1 of 1

Date modified: