Statistical Models for Time Series of Circular Data

数据可以表示为圆圈上的一组观测数据，这些数据产生于一些应用领域，如气象学和生物学。有时，循环观测出现在时间序列的背景下。例如，在气象站每小时测量的一系列风向可以被认为是循环数据的时间序列的例子。另一个例子是在生物学中观察到的候鸟运动方向的时间序列。对于这类数据的建模，我们提出了一些统计模型，并研究了它们的性质。给出了模型的应用实例。结果表明，所提出的模型对某些数据的拟合度要好于现有的熟知过程。

Data which can be expressed as sets of observations on the circle arise in a number of areas of applications such as meteorology and biology. On occasions, circular observations appear in a time series context. For instance, a series of wind directions measured hourly at a weather station can be considered an example of time series of circular data. Another example is a time series of directions of movements of migratory birds observed in biology. For the modeling of such data, we have proposed some statistical models and investigated their properties. Applications of the models have been given. It has been shown that the presented models provide better fits to some data than the well-known existing processes.