Journal: Frontiers in Network Physiology
Article Title: A guide to Whittle maximum likelihood estimator in MATLAB
doi: 10.3389/fnetp.2023.1204757
Figure Lengend Snippet: Periodogram of choleskyfgn (A) , arfima0d0 (B) , whitenoise (C) , and empirical (D) signals with the theoretical power spectral density of fGn (orange curve) and ARFIMA (0, d ,0) (yellow curve). The theoretical power spectral densities were computed with the estimated values of H and d obtained via whittle.m. Those values, entered in MATLAB code 2 and 3, are presented in .
Article Snippet: Then paste the MATLAB codes in the following order: • MATLAB code 1 : Periodogram estimation • MATLAB code 8 : Optimization for fGn-based Whittle’s log-likelihood function • MATLAB code 10 : If the observation vector is non-stationary, fGn-based Whittle’s likelihood • MATLAB code 9 : Optimization for ARFIMA (0, d ,0)-based Whittle’s log-likelihood function • MATLAB code 11 : If the observation vector is non-stationary, ARFIMA-based Whittle’s likelihood • MATLAB code 6 : Whittle’s log-likelihood MATLAB function with fGn theoretical PSD • MATLAB code 7 : Whittle’s log-likelihood MATLAB function with ARFIMA (0, d ,0) theoretical PSD
Techniques: