能够实现SSA分解源码
SSA generates a trayectory matrix X from the original series x1
% by sliding a window of length L. The trayectory matrix is aproximated
% using Singular Value Decomposition. The last step reconstructs
% the series from the aproximated trayectory matrix. The SSA applications
% include smoothing, filtering, and trend extraction.
% The algorithm used is described in detail in: Golyandina, N., Nekrutkin,
% V., Zhigljavsky, A., 2001. Analisys of Time Series Structure - SSA and
% Related Techniques. Chapman & Hall/CR.