Parameter Estimation, Forecasting, Gap Filling
Chapter 3 is devoted to applications of SSA for one-dimensional series for forecasting, gap filling, low-rank approximation, parameter estimation, and change-point detection. The SSA analysis of time series of Chap. 2 is model-free. Methods of Chap. 3, on the contrary, are model-based. The model is constructed on the base of the approximating subspace built in the process of performing the SSA analysis of Chap. 2. The main parametric model is a linear recurrence relation which the signal should approximately satisfy. Application of methods is illustrated on real-life data.
- Basseville M, Nikiforov IV, et al (1993) Detection of abrupt changes: theory and application, vol 104. Prentice Hall, Englewood CliffsGoogle Scholar
- Gillard J, Zhigljavsky A (2011) Analysis of structured low rank approximation as an optimization problem. Informatica 22(4):489–505Google Scholar
- Hyndman RJ (2017) forecast: Forecasting functions for time series and linear models. URL http://CRAN.R-project.org/package=forecast, R package version 8.1, with contributions from Slava Razbash and Drew Schmidt
- Zhigljavsky A, Golyandina N, Gillard J (2016a) Analysis and design in the problem of vector deconvolution. In: Kunert J, Müller HC, Atkinson CA (eds) mODa 11 - advances in model-oriented design and analysis. Springer International Publishing, pp 243–251Google Scholar