Time series analysis and forecast, Extraction of periodicities with varying amplitudes

Examples: Time series analysis

Extraction of periodicities with varying amplitudes

The ‘Caterpillar’-SSA technique allows to extract from the time series the periodic components with varying amplitudes. Natural decomposition of the time series is constructed on the base of the unique parameter (the window length). Grouping paired spiral components of the decomposition, you can extract amplitude-modulated periodics.

Figure 1.11: Drunkenness: varying amplitudes

Data: "Drunkenness"
Description: monthly public drunkenness intakes, Minneapolis, U.S., from January 1966 to July 1978
Source: Time Series Data Library maintained by Rob Hyndman
Parameters: window length = 60
Figure 1.11: the time series component reconstructed from the eigentriples with numbers 4-5 (red line) and the initial time series

Previous example | Next example