Robust Analysis of Time Series

Singular Spectrum Analysis (SSA) is a powerful and widely used non-parametric method to analyze and forecast time series. Although SSA has proven to outperform traditional parametric methods, one of the steps of the SSA algorithm is the singular value decomposition (SVD) of the trajectory matrix, which is very sensitive to the presence of outliers because it uses the 𝐿۲ norm optimization. The main aim of this paper is to introduce two robust alternatives to the SSA. The proposed robust SSA alternatives are compared with the SSA and one available robust SSA algorithm, in terms of model fit via Monte Carlo simulations, considering several contamination scenarios

Keywords:

Singular value decomposition , singular spectrum analysis , time series , robust statistics , contaminated data.