Comparative Assessment of Hierarchical Clustering Methods for Grouping in Singular Spectrum Analysis

Singular spectrum analysis (SSA) is a popular filtering and forecasting method that used in wide range of fields such as time series signal processing. A commonly approach to identify the meaningful components grouping step SSA utilization visual information eigentriples. Another supplementary employing an algorithm performs clustering based on dissimilarity matrix defined by weighted correlation between series. The literature search revealed no investigation has compared various methods. aim this paper was compare effectiveness different hierarchical linkages appropriate groups SSA. comparison performed corrected Rand (CR) index criterion utilizes simulated It also demonstrated via two real-world how one can proceed, step-by-step, conduct using method. This supplemented with accompanying R codes.