Higher-Order Singular Spectrum Analysis for Multichannel Biomedical Signal Analysis

Singular spectrum analysis (SSA) is a nonparametric spectral estimation method that decomposes a time series signal into interpretable components. In this paper, we introduce a new multilinear generalization of SSA, called higher-order multivariate SSA (HO-MSSA), specifically designed to handle multivariate and multichannel time-series signals. The proposed method employs time-delay embedding and tensor SVD to transform multichannel time-series signals into trajectory tensors, subsequently decomposing them into distinct components. Similar to the classical SSA, these components hold significant interpretability. The effectiveness of HO-MSSA is illustrated through its application to the analysis of multichannel biomedical signals.