The fast Fourier transform for experimentalists, part IV: autoregressive spectral analysis

We consider two additional kinds of spectrum estimates: autoregressive (AR) estimates and the maximum entropy (ME) method. In the first approach, we assume that an AR process generates the time series, which means we can compute the PSD of the time series from estimates of the AR parameters. The second approach is a special case of the first, but it uses a different method for estimating the AR parameters. Specifically, it chooses them to make the PSD's inverse transform compatible with the measured time series, while remaining maximally noncommittal about the data outside the observational window.