Course Schedule:

Week Content

 Week 1: Aug 27  

 Week 1: Aug 29

 Week 1: Aug 31

 Mon. Lecture 1: Introduction & Course Overview. (PDF) (R Markdown) (Key)

 Wed. Lecture 2: Time Series Intro & R Basics. (PDF) (R Markdown) (Key)

 Fri. Lecture 2: Time Series Intro & R Basics, cont..

 Week 2: Sept 3  

 Week 2: Sept 5

 Week 2: Sept 7

 Mon. No Class (Labor Day)

 Wed. Lab 1: Time Series Intro & R Basics. (HTML) (R Markdown)

 Fri. Quiz 1, Lecture 3: Time Series Decomposition. (PDF) (R Markdown) (Key)

 Week 3: Sept 10  

 Week 3: Sept 12

 Week 3: Sept 14

 Mon. Quiz 2, Lecture 4: Autocorrelation (PDF) (R Markdown) (Key)

 Wed. Lecture 4: Autocorrelation, cont...,

 Fri. Lab 2: Decomposition and Forecasting (HTML) (R Markdown)

 HW 1 due (HTML) (R Markdown)

 Week 4: Sept 17  

 Week 4: Sept 19

 Week 4: Sept 21

 Mon. Quiz 3,  Lecture 5: Forecasting Intro (PDF) (R Markdown) (Key)

 Wed. Lecture 6: Smoothing (PDF) (R Markdown) (Key)

 Fri. No class: MT ASA Chapter Meeting

 HW 2 due (HTML) (R Markdown)

 Week 5: Sept 24

 

 Week 5: Sept 26

 Week 5: Sept 28 

 Mon. Quiz 4, Lecture 5: Smoothing, cont...

  Lab 3: Forecasting (HTML) (R Markdown)

 Wed. Lecture 7: Stochastic Models Part 1 (PDF) (R Markdown) (Key)

 Fri. Lecture 8: Stochastic Models Part 2 (PDF) (R Markdown) (Key)

 HW 3 due (HTML) (R Markdown)

 Week 6: Oct 1

 Week 6: Oct 3

 

 Week 6: Oct 5

 Mon. Quiz 5, Stochastic Models Part 2

 Wed. Lab 4: AR processes (HTML) (R Markdown)

  Lab 4, Part 2 (536): Recursive Bayesian estimation (HTML) (R Markdown)

 Fri. Lecture 9: Regression Part 1 (PDF) (R Markdown) (Key)

 HW 4 due (HTML) (R Markdown)

 Week 7: Oct 8

 Week 7: Oct 10

 Week 7: Oct 12

 Mon. Quiz 6. Lecture 9: Regression Part 1, cont...

 Wed. Lecture 10: Regression Harmonics (PDF) (R Markdown) (Key)

 Fri. Lab 5: Regression (HTML) (R Markdown)

 HW 5 due (HTML) (R Markdown)

 Week 8: Oct 15

 Week 8: Oct 17

 Week 8: Oct 19

 Mon. Quiz 7,  Lab 6: Regression with Uber Pickups (HTML) (R Markdown), Lab 6, Part 2 (536): Recursive Bayes for Moving Target (HTML) (R Markdown)

 Wed. Lecture 11: Regression and Forecasting (PDF) (R Markdown) (Key)

 Fri.  Lab 7: Regression and Forecasting (HTML) (R Markdown)

 Week 9: Oct 22

 Week 9: Oct 24

 Week 9: Oct 26 

 Mon. Lab 5, Lab 6, Lab 7 due.  Review Session

 Wed.  In class exam

 Fri.  536 Only - State Space Models Lecture and Course Projects (PDF)

 Week 10: Oct 29

 

 Week 10: Oct 31

 Week 10: Nov 2

 Mon. Take Home Exam Due.

 Lecture 12: Moving Average Processes (PDF) (R Markdown)

 Wed. Lecture 12: Moving Average Processes (PDF) (R Markdown) (Key)

 Fri.  Lecture 13: ARMA models (PDF) (R Markdown) (Key)

 HW 6 due (536 only) (HTML) (R Markdown)

 Week 11: Nov 5

 Week 11: Nov 7

 Week 11: Nov 9 

 Mon. Quiz 8. Lab 8: MA and ARMA models (HTML) (R Markdown)

 Wed. Lecture 14: State Space Models & Exponential Smoothing (PDF) (R Markdown) (Key)

 Fri. Lecture 15:  State Space Models & ARMA Models (PDF) (R Markdown) (Key)

 Week 12: Nov 12

 

 Week 12: Nov 14

 Week 12: Nov 16 

 Mon. No Class (Veteran's Day) 

  HW 7 due (HTML) (R Markdown)

 Wed. Quiz 9, Lecture 16: ARIMA models (PDF) (R Markdown) (Key)

 Fri. Lab 9: ARIMA models (HTML) (R Markdown)

 Week 13: Nov 19

 

 Week 13: Nov 21

 Week 13: Nov 23

 Mon. Demo: Dynamic Regression Models (HTML) (R Markdown)

  HW 8 due (HTML) (R Markdown)

 Wed. No Class (Thanksgiving Break)

 Fri. No Class (Thanksgiving Break)

 Week 14: Nov 26

 Week 14: Nov 28

 Week 14: Nov 30

 Mon. Quiz 10, Lecture 17: Multivariate Time Series (PDF) (R Markdown) (Key)

 Wed. Lab 10: Multivariate Time Series (HTML) (R Markdown)

 Fri. Course Review / Putting it all together

 Week 15: Dec 3

 Week 15: Dec 5

 Week 15: Dec 7 

 Mon. In class final exam

 Wed. Lecture 18: Heteroskedastic Variance Models (PDF) (R Markdown) (Key)

 Fri. Lab 11: Heteroskedastic Variance Models (HTML) (R Markdown)

  Finals Week:

Dec 9: 9:00 AM

Dec 10: 8 - 9:50 AM

 

Take home exam due

STAT 536 Project Presentations. Project Summary (PDF)

STAT 436/536 Overview:

  • Meeting Time: Monday, Wednesday, Friday -  12:00 - 12:50 
  • Classroom: Wilson Hall 1-147
  • Office Hours: Monday/Wednesday 2  -  3 or by appointment

Course Description

This course introduces the statistical methodology and models to analyze time series data with an emphasis on both the theory of time series covariance structures and the practical implementation of time series modeling using R.

STAT 436 and STAT 536 will be taught concurrently in the same room; however, there will be times when the different sections are working on different activities. Unlike STAT 511 or STAT 512, STAT 536 will be a mathematically rigorous course targeting M.S. and Ph.D. statistics students.

Prerequisites

One of: STAT 411 / STAT 511, consent of instructor. For STAT 536: a graduate level probability and inference sequence (STAT 505/506), linear models background (STAT 501/502), and experience with R is strongly suggested.

Course Objectives

At the completion of this course, students will be able to:

  1. Identify trends and seasonal effects in time series data,
  2. Use statistical software to fit ARIMA model to address correlation in time series data,
  3. Understand and describe uncertainty in time series forecasting, and
  4. Summarize inferences from times series analyses orally and in writing.

Textbook:

  1. Introductory Time Series with R, by Paul Cowpertwair and Andrew Metcalfe
  2. Dynamic Linear Models with R, by Giovanni Petris, Sonia Petrone, and Patrizia Campagnoli (Optional for 436, will be used for 536)

Both textbooks are from the Springer series and electronic versions are available through the Montana State Library website.

Course Evaluation:

  • Quizzes: 10% of final grade:

    • There is no formal attendance policy, but there will be weekly quizzes. Typically quizzes will be posted on D2L in advance of class, but may also be given in class. 

  • Homework: 25% (436) / 20% (536) of final grade
    • Homework problems will be assigned every week. Students are allowed and encouraged to work with classmates on homework assignments, but each student is required to write their own homework.
  • Midterm Exam 20% & Final Exam 20%  of final grade
    • Exams will have two components: an in-class exam and a take home portion. The in-class portions will be largely conceptual including some short mathematical derivations. The take home portions will focus on the analysis of data and implementation of time series methods.
    • All take home exams are to be completed strictly on an individual basis.
  • Lab 25% (436) / 10% (536) of final grade
    • The course will periodically have labs, which will be organized group activities to be completed in class. Typically, different labs will be assigned for 436 and 536.
    • Students should plan to bring computers to class each day to be prepared for labs and active learning exercises in class.
  • Project 20% (536) of final grade

    •  The project will only be for students enrolled in STAT 536.

    • The final project will focus on the complete data analysis cycle for time series models using relevant data selected by the student and agreed upon by the instructor and student.