The course starts with a lecture introducing time series modeling strategy and its usage. Then, several lectures are devoted to Box-Jenkins modeling procedures for univariate time analysis, including identification, estimation, diagnostic checking and forecasting of a time series model. Seasonal models with applications is also discussed. Two important topics, intervention and outlier, are illustrated with examples followed by some discussions on unit root test. Finally, the class ends with some discussions on the analysis of public health data.
Similar to univariate time series, multiple time series modeling procedure consists of identification, estimation, diagnostic checking and forecasting. Methods with applications will be discussed.
One can not really master time series models without actually using them for real data analysis. Scientific Computation Association (SCA) is the statistical package adopted in this course. Students are required to use SCA for homework and a term project.