Singular spectral analysis‐based denoising without computing singular values via augmented Lagrange multiplier algorithm

This study proposes an augmented Lagrange multiplier-based method to perform the singular spectral analysis-based denoising without computing the singular values. In particular, the one-dimensional (1D) signal is first mapped to a trajectory matrix using the window length null L . Second, the trajectory matrix is represented as the sum of the signal dominant matrix and the noise-dominant matrix. The determination of these two matrices is...