Syllabus
I. Classification of random processes.
II. Decomposition methods: 1. Trend. 2. Seasonality and periodicity. 3. Tests of randomness.
III. Box-Jenkins methodology 1. ARMA models ARMA 2. Identification, estimation, verification and prediction. 3. ARIMA and seasonal models.
IV. Financial time series: 1. Models of volatility (GARCH). 2. Models nonlinear in mean.
V. Multivariate time series (vector autoregression, Kalman filter).
Annotation
Basic methods of time series analysis and their applications, time series decomposition and adaptive techniques,
Box-Jenkins methodology including ARIMA and seasonal models, financial time series (models of volatility and nonlinear in mean), multivariate time series (vector autoregression, Kalman filter). Most of the methods are applied in a facultative seminar. Requirements: Basic knowledge of statistics.