Nonlinear prediction ladder-filters for higher-order stochastic sequences (Q1059042)

These notes deal with the problem of nonlinear least squares prediction of higher-order stochastic sequences using nonlinear orthogonal digital filters, and presents efficient algorithms for computation of filters of the Volterra-Wiener class. Chapter 1 introduces the subject, chapter 2 introduces the nonlinear least sequences prediction problem algebraically and geometrically, chapter 3 presents recursive geometric solutions to the nonstationary nonlinear prediction problem and chapter 4 is devoted to the problem of complexity induction in the nonlinear ladder-filter.