Statistics 520
Time Series and Applications
Spring 2021
Instructor: Chong Gu
Classes: 1:30  2:20 MWF, WTHR 160
Office Hours: 11:45  1:15 MF, or by appointment, at
https://purdue.webex.com/meet/chong.
Office Hour on Friday 4/2/21 will start at 12:10pm.
The midterms will be on Wednesdays, 3/3 and 4/21, at 810pm in LWSN B151; you can bring 4 pages of crib sheets, lettersize doublesided. Old exams will be posted at appropriate times.
An old exam can be found here, showing you the format of the exam.
 Course outline
 This course offers an introduction to the analysis of time
series. Topics to be covered include the autocorrelation and
spectrum of stationary processes, the structure, estimation, and
identification of AutoRegressive (Iterated) Moving Average
(ARIMA) models, forecasting, model diagnostics, seasonal models,
and transfer function models. Software tools will be an
important part of the course, for which we will mainly draw on
the resources available in R, an opensource programming
environment for data analysis and graphics.
 Prerequisite
 Basic concepts of probability theory, working knowledge of
statistical inference, linear models, and matrix algebra.
 Textbook

 Time Series Analysis: Forecasting and Control, by Box, Jenkins, and Reinsel.
The data sets used in the text are given here.
 References

 Software
 We will be using R, an opensource programming environment for
data analysis and graphics, as the primary platform for
computation and graphics. R resources are to be found at The Comprehensive R Archive
Network.
 Course Work
 There will be about 78 assignments, with "written" and "lab"
problems mixed in; the assignments are due at Brightspace under
assignments. There will also be two inperson
written midterms and a final lab project. The midterms are
closedbook but you are allowed 4 pages of lettersize
doublesided crib sheets.
 Grading
 The letter grade will be based on assignments (30%), midterms
(2 x 30%), and the final project (10%).
 Lecture Notes
 Assignments

