Functional singular spectrum analysis

In this paper, we develop a new extension of the singular spectrum analysis (SSA) called functional SSA to analyze time series. The methodology is constructed by integrating ideas from data and univariate SSA. Specifically, introduce trajectory operator in world, which equivalent matrix regular SSA, one needs obtain value decomposition (SVD) decompose given Since there no procedure extract SVD (fSVD) operator, computationally tractable algorithm fSVD components. effectiveness proposed approach illustrated an interesting example remote sensing data. Also, efficient user‐friendly R package shiny web application allow interactive exploration results. (Less)