Accession Number : ADA133945

Title :   Study on Extremizing Adaptive Systems and Applications to Synthetic Aperture Radars.

Descriptive Note : Annual scientific rept. 10 Sep 82-9 Sep 83,

Corporate Author : ENVIRONMENTAL RESEARCH INST OF MICHIGAN ANN ARBOR

Personal Author(s) : Politis,Demetrios

PDF Url : ADA133945

Report Date : May 1983

Pagination or Media Count : 6

Abstract : Klopf's work on the functioning of the neuron was studied and critically examined for engineering application possibilities. Similarly, Barto's work on the implementation of Klopf's ideas in computer simulated nets/systems was studied to determine if it could provide suitable models for physical systems. The latest learning system investigated by Barto, described as 'Learning with an Adaptive Critic' was considered as the most promising for engineering applications. A functional engineering model of that system has been developed and its dynamic behavior of this system is currently being investigated in order to improve our understanding of the system operation and potential applications. In parallel with this study we are looking for possible application of such learning systems in synthetic aperture radars and data exploitation. Several potential applications have already been suggested. These suggestions will be further explored and the most promising will be proposed for full investigation and possible implementation.

Descriptors :   *Adaptive systems, Bionics, Nerve cells, Computerized simulation, Learning, Searching, Feedback, Synthetic aperture radar, Neural nets, Sampling, Weighting functions, Models

Subject Categories : Active & Passive Radar Detection & Equipment
      Bionics

Distribution Statement : APPROVED FOR PUBLIC RELEASE