STAT 804
Time Series Analysis

Instructor: Richard A. Lockhart
Office: TLX 10546        Phone: 3264        Messages: 3331         e-mail:

This course is intended to survey both time-domain and frequency domain analysis of time series. I expect you all to be familiar with the basics of the multivariate normal distribution and complex arithmetic. I will develop Fourier methods briefly where necessary.

There will be 2 two hour lectures per week. I will actually lecture for 100 minutes then have open discussion.

Recommended Textbook:

Box, G. E. P. , Jenkins, G. M.  and Reinsel, G. (1994) Time series analysis : forecasting and control , 3rd ed. Englewood Cliffs: Prentice Hall. I hope my notes together with references will be reasonably self-contained so that possession of Box, Jenkins and Reinsel is not precisely necessary; the text is, however, an absolutely standard reference in the field.

Other references

Brillinger, D. R. (1975) Time Series: Data Analysis and Theory. New York: Holt, Rinehart and Winston.

Brockwell, P. J. and Davis, R. A. (1991) Time Series: Theory and Methods, 2nd ed. New York: Springer-Verlag.

Priestley, M. B. (1981) Spectral Analysis and Time Series. London: Academic Press.

Shumway, Robert H. (1988) Applied Statistical Time Series Analysis. Englewood Cliffs: Prentice Hall.

The course will be graded on the basis of assignments together with a take home final exam. You will do a good deal of computing in S using the package on the math Unix network. If you do not have an account yet you must leave me a note with your full name and, for students from outside Mathematics and Statistics, your campus address. I will expect you to produce software in S; except for an introductory session in the first or second week I will expect you to learn this package on your own.

Richard Lockhart