Title :
Basis Functions of the Multidimensional CMAC Neural Network
Author_Institution :
Dept. of Radio Eng. & Cybern., Moscow Inst. of Phys. & Technol. (State Univ.), Moscow, Russia
Abstract :
The short introduction to the cerebellar model articulation controller (the CMAC neural network) and its main properties are presented. It is shown that the basis functions of the CMAC neural network are given implicitly, the number of these function is finite, and they have the incorrect pyramid forms. The expressions for the number of these function is given.
Keywords :
cerebellar model arithmetic computers; learning (artificial intelligence); CMAC neural network; basis function; cerebellar model articulation controller; learning algorithm; Accuracy; Approximation methods; Artificial neural networks; Mathematical model; Noise; Training; CMAC neural network; basis functions; learning algorithm;
Conference_Titel :
Engineering and Telecommunication (EnT), 2014 International Conference on
Print_ISBN :
978-1-4799-7011-7
DOI :
10.1109/EnT.2014.23
