A design of adaptive stack filters using pattern classification

This paper proposes a novel design technique of adaptive stack filters. The Boolean function of the proposed filters is selected by a controller in which pattern classifier is introduced. This controller classifies an input vector into some cluster according to its pattern and selects the Boolean function corresponding to its cluster. The Boolean functions used in this technique are adjusted by using adaptive least mean absolute (ALMA) algorithm. Finally, a simulation example is given to illustrate the utility of the proposed technique.