Business 41910: Time Series Analysis for Forecasting and Model Building
Quarter of 2008
Instructor: Ruey S. Tsay
Office: HPC 455
Class: Thursday 8:30 am -- 11:30 am,
starting September 25.
(No class in the Thanksgiving week.)
Office hour: (a) Wednesday, 11:00 am --
(b) By appointment
You may e-mail me questions. E-mail is the easiest way
to make contact with me. I try to check the e-mail at least
once a day.
Mr. Paco Vazquez-Grande. Email: email@example.com
No textbook is used
(a) Time Series Analysis: Forecasting and Control. Box, Jenkins and Reinsel (2008), 4rd Ed. Wiley [First 9 chapters]
(b) Analysis of Financial Time Series, 2nd Edition, Tsay (2005), Wiley.
[Chapters 2, 11, and 12]
(c) A Course in Time Series Analysis: Pena, Tiao and Tsay (2001), Wiley
(d) Time Series Analysis by State Space Methods: Durbin and Koopman (2001),
Oxford University Press.
(e) Time Series Analysis: Hamilton (1994), Princeton University Press.
(f) Additional books are given in the course syllabus.
Some reference articles will also be given.
Midterm 30% + Final Exam 40% + Homework 30%
where scores of each component are normalized to be out of 100.
No late homework assigments are accepted.
Week 1: Introduction, difference equations and linear models: lec1-08.pdf
data sets used: dexuseu.txt, gdp.txt, m2ns.txt, payems.txt, ppiaco.txt, unrate.txt
Week 2: Unit-root, ARIMA models, seasonal models, and aggregation:
lec3-08.pdf, lec4-08.pdf, and lec5-08.pdf
Week 3: Forecasting and model building: lec6-08.pdf, lec7-08.pdf
Week 4: Model specification and Estimation: lec8-08.pdf , lec9-08.pdf
Data set used: airpml.txt
Week 5: lec10-08.pdf
Week 6: Unit root and unit-root test: lec11-08.pdf data: vix08.txt
Week 7: State-Space Models and Kalman Filter: lec12-08.pdf
Week 8: MCMC Methods and Their Applications: lec13-08.pdf
Week 9: Conditional Heteroscedasticity: lec14-08.pdf
Week 10: Duration Models: lec15-08.pdf (Note the time & room changes:
December 3, 1:30 pm to 4:30 pm at C24.)
(1) R and SCA will be used for most analyses in class. SCA is available on
all GSB machines. For R, the recent version 2.6 or higher is recommended.
(2) S-Plus is also used mainly for Kalman filtering.
(3) Students may use other packages, e.g., SAS, SPSS, MATLAB,
to answer hw questions.
commands for SCA and R:
R code for EACF: eacfR.txt
R code for out-of-sample forecast: r-backtest.txt
R code for out-of-sample forecast error transform: r-foreden.txt
Week 1: HW1 (due in one week; before the lecture). Solution.
Week 2: HW2 (due on October 9 before class). Solution.
Week 3: HW3 (due on October 16 before class). Data sets used:
bjsera.txt, dat4a.txt, dat4b.txt, dat4c.txt, dat4d.txt, dat4e.txt Solution
Week 4: HW4 (due on October 23 before class).
Week 7: HW5 (due on November 13 before class). Data sets:
m-ppiaco08.txt, w-m2.txt, ginterv.txt, m-cpiausl.txt
Week 8: HW6 (due on November 20 before class). Solution
Midterm: Week 6, October 30, 9:30 am - 11:30 am (Lecture before the test)
Open books and notes. Solutions: exam08s
Exam: December 11, 8:00 am to 11:00 am
Solutions to Final Exam: final08s.pdf