Welcome to CSI 678 / STAT 658
Time Series Analysis and Forecasting
Fall, 2012
Instructor:
James Gentle
Lectures: Thursdays 4:30pm  7:10pm, Innovation Hall 133
If you send email to the instructor,
please put "CSI 678" or "STAT 658" in the subject line.
Course Description
Time series analysis is used for diverse applications in economics,
the social sciences, the physical and environmental sciences, medicine,
and signal processing. This course presents the fundamental principles of
time series analysis including mathematical modeling of time
series data and methods for statistical inference. Models in both the
time domain and the frequency domain will be developed and explored.
The emphasis, however, will be on the time domain.
An integral part of
the course is the use of R for simulation, calculation, and implementation of
time series analysis techniques. Students may use other software for
assignments if they are familiar with the software.
Prerequisites
The prerequisites for this course include STAT 544 or ECE 528 or an equivalent
multivariable calculusbased graduate course in Applied Probability or
Random Processes.
Text and Software
The text is
Time Series Analysis and Its Applications: With R Examples
third edition, 2010, by Robert H. Shumway and David S. Stoffer.
ISBN # 9781441978646
Authors' website for text.
The software used in this course is R, which is a freeware package that can be
downloaded from the
Comprehensive R Archive Network (CRAN).
It is also available on various GMU computers in student labs.
No prior experience in R is assumed for this course.
A good site for getting started with R, especially for people who are somewhat
familiar with SAS or SPSS, is
Quick R.
The text provides examples in R, and the
associated website provides data files and R scripts that can be downloaded.
The authors have also provided an R package, astsa
. In
many assignments, the R function library(astsa)
will be used
to load the package, along with its builtin datasets.
If students are familiar with other software, they may find it instructive to
translate the R scripts into the other language.
Students may use other software for assignments, but use of other software will
not be covered in the course, and the data used in the text will have to be
obtained from portable files, rather than as R datasets, which are available
directly from the authors' package.
Grading
Student work in the course (and the relative weighting of this work
in the overall grade) will consist of
homework assignments (30)
midterm exam (30)
final exam (40)
Homework
Each homework will be graded based on 100 points, and 5 points will be deducted
for each day that the homework is late.
Start each problem on a new sheet of paper and label it clearly.
Homework will not be accepted as computer files (and certainly not as
faxes!); it must be submitted on
paper.
Academic Honor
Each student enrolled in this course must assume the
responsibilities of an active participant in GMU's scholarly
community in which everyone's academic work and behavior are
held to the highest standards of honesty. The GMU policy on
academic conduct will be followed in this course.
Collaborative Work
Students are free to discuss homework problems or other topics
with each other or anyone else, and are
free to use any reference sources. Group work and discussion outside of
class is encouraged, but of course explicit copying of homework solutions
should not be done.
Note that other courses at GMU may have different restrictions on discussions
among students relating to course content.
Schedule
The details of the schedule will evolve as the semester progresses.
Week 1, August 30

Course overview; notation; etc.

The R program.

Plots:

layout(1:3)
sets up to plot 3 graphs in same display window.

lines()
or points()
after doing plot()
adds lines or points to that same plot.
 If
x
is an ordinary numeric vector,
plot(x,type="l",col="red")
produces a red line plot.
 If
y
is an ordinary numeric vector,
points(y,col="blue")
prints blue points on the same plot produced by
the previous statement.

Nature of time series data.

Simulation in R.
Assignment: Read Shumway and Stoffer, Sections 1.1 through 1.6. Look over
Appendix R.
Assignment 1a, due September 6: In Shumway and Stoffer: problems
1.1.
Assignment 1b, due September 13: In Shumway and Stoffer: problems
1.2, 1.3, 1.4, 1.6.
Week 2, September 6

Means, autocovariance, autocorrelation, crosscovariance, crosscorrelation.

Stationarity.

Estimation of time series parameters.

Vectorvalued time series.

Convergence of sequences of random variables (estimators).

Exploratory data analysis and comparison of statistical models.

Regression and time series.

Simulation in R.
Assignment: Read Shumway and Stoffer, Sections 1.7 through 2.1.
Assignment 2, due September 13: In Shumway and Stoffer: problems
1.8, 1.13, 1.19, 1.22, 1.25, 2.1.
Week 3, September 13
R functions for datahandling.
Smoothing time series.
Gaussian regression models.
Modified
Assignment 3, due September 20:
In Shumway and Stoffer Third Edition: problems
2.6, 2.11(a),(b),(c).
All problem numbers henceforth will refer to the problems in the Third Edition of
the text.
Week 4, September 20
Differencing, detrending, etc.
More on exploratory data analysis, regression, smoothing,
and comparison of statistical models.
Discuss problems from the text.
Modified
Assignment 4, due October 11: In Shumway and Stoffer: problems
2.2, 2.8, 2.9.
Week 5, September 27
Discuss homework and any miscellaneous questions.
Use of sample ACF and CCF for statistical inference.
Example 2.9: a quick introduction to the frequency domain.
Assignment 5, due October 11: In Shumway and Stoffer: problem
1.21 with the addition:
use your estimated value of the ACF to set a 95% twosided confidence interval
for the ACF, using Property 1.1 on page 29.
Week 6, October 4
Inclass midterm exam.
Closed book and closed notes.
Covers material in Shumway and Stoffer through approximately
the end of Chapter 2, except for the properties of matrices in least
squares analysis, the periodogram, and the
asymptotic properties of sample statistics.
Week 7, October 11
Sample ACF, CCF, and their largesample properties.
ARMA models; basic setup and notation.
Difference equations.
Causality and invertibility
Assignment 6, due October 18: In Shumway and Stoffer: problems
3.1, 3.2(a) (b) (c) (d) (e), 3.4(a) (b), 3.5, 3.7(a) (b).
Week 8, October 18
More on ARMA models (Sections 3.4, 3.5, 3.6).
ACF and PACF.
Estimation.
Assignment 7, due November 1: In Shumway and Stoffer: problems
3.8, 3.9, 3.10, 3.11, 3.16.
Week 9, October 25
Interesting and useful sources of data:
General data assembled by the St. Louis Fed:
research.stlouisfed.org/fred2/.
Other data from the Fed Board of Governors:
www.federalreserve.gov/econresdata/statisticsdata.htm.
Interest rate data from the Fed Board of Governors:
www.federalreserve.gov/econresdata/statisticsdata.htm.
Other data from the Fed Board of Governors:
www.federalreserve.gov/releases/h15/update/.
Data from the World Bank:
data.worldbank.org/.
Stock price data:
finance.yahoo.com/.
Get data as .csv file; read in R by read.csv.
Can read data directly from web.
There is a useful R program written by John Nolan
that is especially useful for reading and manipulating stock prices from Yahoo.
Forecasting using ARMA models.
More on ARMA models and ARIMA models (Sections 3.6, 3.7, 3.8, 3.9).
Assignment 8, due November 1: Download the quarterly US GDP in
chained billions of dollars (to 1 decimal place) from 19470101 to
20120701. (You can obtain it from the St. Louis Fed's FRED site.)
(a) Make an R ts object to contain the data. (Frequency =?).
(b) Plot the raw data and comment on any obvious trends or seasonalities.
(c) Attempt to remove any trend by differencing. Plot and comment.
(d) As with much economic data, diff log may be a useful transformation.
Make this transformation and plot the transformed data.
(e) Compute the sample ACF and PACF.
(f) Fit an AR(p) to the diff logs (what p did you choose, and why?).
Obtain the residuals and make a QQ plot
against a normal distribution (use qqnorm). Comment.
(g) Fit an MA(q) to the diff logs (what q did you choose, and why?).
Obtain the residuals and make a QQ plot
against a normal distribution. Comment.
(h) Fit an ARIMA(p,d,q) model, and discuss.
(This assignment is similar to Example 3.38 in Shumway and Stoffer.)
Week 10, November 1
More on forecasting.
Seasonal models.
General review of ARIMA models.
Comments.
Assignment 9, due November 8: In Shumway and Stoffer: problems
3.19, 3.28, 3.33, 3.35, 3.36.
Week 11, November 8
The periodogram and spectral density.
Assignment 10, due November 15: In Shumway and Stoffer: problems
4.1, 4.4, 4.5, 4.6.
Week 12, November 15
The periodogram and the discrete Fourier transform.
Nonparametric spectral estimation.
Cross spectra.
Assignment 11, due November 29: In Shumway and Stoffer: problems
4.8, 4.9, 4.12, 4.13.
November 22:
No class.
Week 13, November 29
More on the periodogram and nonparametric spectral estimation.
Linear filters.
Assignment 12, due December 6
(but will not be collected).
In Shumway and Stoffer: problems 4.18, 4.19, 4.27, 4.29.
Week 14, December 6
Additional topics (slide 23 and following, from the
leture notes for week 13.)
Parametric spectral estimation.
Wavelets.
Lagged regression.
Optimal signal extraction.
General review.
Comments.
December 13
4:30pm  7:15pm Final Exam.
Closed book and closed notes.