Information Theoretic Results for Stationary Time Series and the Gaussian-Generalized von Mises Time Series

AbstractThis chapter presents some novel information theoretic results for the analysis of stationary time series in frequency domain. In particular, spectral distribution that corresponds to most uncertain or unpredictable with values autocovariance function fixed is generalized von Mises distribution. It thus a maximum entropy and corresponding called series. The used directional statistics modeling planar directions follow multimodal Furthermore, Gaussian-generalized times presented as maximizes entropies domains, respectively, referred temporal entropies. Parameter estimation computational aspects this are briefly analyzed.