Generalized adaptive neural filters and filters based on neural network

Бесплатный доступ

The structure and properties of nonlinear adaptive filter group (NAF) are considered. The filters are based on single neurons and on neural networks. The precision parameters - the upper level and mean absolute error of filters - are analyzed. Such NAF extend the class of stack filters and have better performance in noise suppression.

Filters, neural network, precision, signal, noise, adaptation, performance, quality

Короткий адрес: https://sciup.org/14264745

IDR: 14264745

Статья научная