A systematic presentation of many statistical techniques for the
analysis of time series data. The core topics include time
dependence and randomness, trend, seasonality and error,
stationary processes, ARMA and ARIMA processes, multivariate time
series models and state-space models.
An additional topic is forecasting.
The emphasis is on the theory and methodology of the time-domain
analysis based on ARIMA and state-space models.
An important component of
the course is that the analysis of data sets is illustrated throughout.
interpretation of the results obtained on some assigned problems.
The text is P.J. Brockwell and R.A. Davis,
Introduction to Time Series
and Forecasting, Springer--Verlag. Some materials from P.J. Brockwell
R.A. Davis, Time Series: Theory and Methods, Springer--Verlag, are
imported as necessary.
The course will be evaluated by assignments, one midterm, and a final
Prerequisite:AS/SC/MATH 3033 3.0 and AS/SC/AK/
MATH 3131 3.0, or permission of the course coordinator.
Exclusions:SC/AS/COSC 4242 3.0, SC/EATS
4020 3.0, AS/SC/MATH 4830 3.0, AS/SC/MATH 4930C 3.0,
SC/PHYS 4060 3.0, SC/PHYS 4250 3.0.
Coordinator: P. Song