Singular spectrum analysis (SSA) is a non-parametric spectral decomposition technique for time series, akin to fourier or wavelet analysis, in which a time-series is decomposed into a time-frequency matrix. However, SSA does not rely on strict parametric forms and is able to pull out non-stationary and complex components from time-series in a data-dependent manner. Please refer to the documentation in the methods of SSA.m for details.