Nonlinear foltering of time series using the wavelet transform
В. И. Лобач · Digital Library of the Belarusian State University (Belarusian State University) · 1995
A fast nonlinear filtering algorithm is presented. This algorithm propagates the entire conditional probability functions recursively in a computationally efficient manner using the discrete wavelet transform. With the multiresolution analysis capability offered by the wavelet transform we can speed up the computation by ignoring the high-frequency details of the probability function up to a certain level.