Title :
On generalized adaptive neural filters
Author :
Zhang, Zeeman Z. ; Ansari, Nirwan ; Lin, Jm-Hsang
Author_Institution :
Dept. of Electr. & Comput. Eng., New Jersey Inst. of Technol., Newark, NJ, USA
Abstract :
The generalized adaptive neural filter (GANF), a new class of nonlinear filters, is introduced. It is effective for non-Gaussian noise suppression. Some properties of GANF are derived, and an algorithm for finding the optimal GANF, based on the upper bound in the minimum absolute error, is proposed. The implementation of the optimal GANF by using the least mean square error and the least perceptron error is also discussed. Experimental results are presented to demonstrate the effectiveness of the new filter
Keywords :
adaptive filters; filtering and prediction theory; neural nets; generalized adaptive neural filters; least mean square error; least perceptron error; minimum absolute error; nonlinear filters; optimal GANF; upper bound; Adaptive filters; Adaptive signal processing; Digital signal processing; Error correction; Nonlinear filters; Process design; Signal design; Signal processing algorithms; Stacking; Upper bound;
Conference_Titel :
Neural Networks, 1992. IJCNN., International Joint Conference on
Conference_Location :
Baltimore, MD
Print_ISBN :
0-7803-0559-0
DOI :
10.1109/IJCNN.1992.227329
