Application of Jordan Recurrence Principle in Rosenblatt`s Perceptron
Автоматика и вычислительная техника 2009
Sergejs Jakovlevs, Arkādijs Borisovs

This paper describes a possibility of combining Rosenblatt’s perceptron with Jordan network. As a result, the perceptron gets the property of recurrence. At the same time, teaching remains as fast and effective as in Rosenblatt perceptron. The recurrent perceptron RP-1 obtained in the above way is tested on the problem of teaching the agent which is placed into the simulation environment.


Keywords
Perceptrons, Rekurentais neironu tīkls

Jakovlevs, S., Borisovs, A. Application of Jordan Recurrence Principle in Rosenblatt`s Perceptron. Автоматика и вычислительная техника, 2009, N 1, pp.46-55. ISSN 0132-4160.

Publication language
Russian (ru)
The Scientific Library of the Riga Technical University.
E-mail: uzzinas@rtu.lv; Phone: +371 28399196