Searching for Signal in Noise by Random-Lag Singular Spectrum Analysis

Singular spectrum analysis, a technique to detect oscillations in short and noisy time series, was Ðrst developed for geophysical applications. This work o†ers a generalization for long and noisy time series in astrophysical applications. The motivating problem is the detection of low-amplitude solar oscillations. Subject headings : methods : data analysis È Sun: oscillations