A0521
Title: Complex valued time series modeling in relation to directional statistics
Authors: Takayuki Shiohama - Nanzan University (Japan) [presenting]
Takayuki Shiohama - Nanzan University (Japan)
Abstract: Stationary time series fluctuation often shows periodic behavior and these patterns are usually summarized via a spectral density. Since the spectral density is a periodic function, it can be modeled by using a circular distribution function. Several time series models are studied in relation to a circular or a cylindrical distribution. First, as an introduction, we illustrate how to model bivariate time series data using complex-valued time series in the context of circular distribution functions. Next, some other time series modeling by incorporating cylindrical distributions is illustrated. The maximum likelihood estimation procedures are introduced to estimate unknown model parameters. Some real data analyses are also performed to illustrate the proposed models' applicability.
