Autors: Manolov, E. D., Lyubenov, P, B. Title: DESIGN AND INVESTIGATION OF TWO-PARAMETER SPACE CLASSIFICATION CIRCUITS USING FPAA Keywords: Parameter space, Classification circuits, Reconfigurable har Abstract: The paper presents the results from the design and investigation of two-parameter space classification neural circuits by using Field Programmable Analog Array (FPAA) of Anadigm Inc. To this aim, the functional model of the artificial neuron is discussed and the results from implementation of FPAA variants of it structure are described. The use of the FPAA ensures possibilities for simple programming and dynamic reconfiguration of different values of the weights on the inputs as well as a flexible realization of the arbitrary output transfer functions. The application of the artificial References
Issue
|
Вид: пленарен доклад в международен форум