Skip to content

Files

Latest commit

ionidesionides
update index again
Mar 24, 2025
43da99f · Mar 24, 2025

History

History
196 lines (107 loc) · 7.8 KB

index.md

File metadata and controls

196 lines (107 loc) · 7.8 KB
title author output
DATASCI/STATS 531/631 (Winter 2025) <br>'Modeling and Analysis of Time Series Data'
Instructor: Edward L. Ionides
html_document
toc
true

Course description

This course gives an introduction to time series analysis using time domain methods and frequency domain methods. The goal is to acquire the theoretical and computational skills required to investigate data collected as a time series. The first half of the course will develop classical time series methodology, including auto-regressive moving average (ARMA) models, regression with ARMA errors, and estimation of the spectral density. The second half of the course will focus on state space model techniques for fitting structured dynamic models to time series data. We will progress from fitting linear, Gaussian dynamic models to fitting nonlinear models for which Monte Carlo methods are required. Examples will be drawn from ecology, economics, epidemiology, finance and elsewhere.

Additional information is in the syllabus. Online discussion is on piazza.

631 includes a reading group where we discuss a research paper each week. Students registering for 631 are expected to have taken at least one core PhD-level class such as STATS 600.

Class notes

  1. Introduction

  2. Estimating trend and autocovariance

  3. Stationarity, white noise, and some basic time series models

  4. Linear time series models and the algebra of ARMA models

  5. Parameter estimation and model identification for ARMA models

  6. Extending the ARMA model: Seasonality, integration and trend

  7. Introduction to time series analysis in the frequency domain

  8. Smoothing in the time and frequency domains

  9. Case study: An association between unemployment and mortality?

  10. Forecasting

  11. Introduction to partially observed Markov process models

  12. Introduction to simulation-based inference for epidemiological dynamics via the pomp R package

  13. Simulation of stochastic dynamic models

  14. Likelihood for POMP models: Theory and practice

  15. Likelihood maximization for POMP models

Homework and participation assignments

Quizzes

Midterm project

Final project

  • You're welcome to browse previous final projects. The 2024, 2022 and 2021 final projects have a posted summary of peer review comments. Earlier projects from 2016, 2018, 2020 may also be useful.

If building on old source code, note that there are some differences between versions of the software package pomp. The pomp version 2 upgrade guide can be helpful. There are various smaller changes between pomp 2.0 and the current pomp.

Using the Great Lakes cluster

  • Great Lakes access will be set up after the midterm project and used for the second half of the course.

  • Introductory notes for using our class account on the greatlakes cluster. This is optional but may be helpful for your final project.

  • If you are already familiar with using R on Great Lakes, all you need to know is the class account: datasci531w25_class.

  • You are expected to use our class account only for computations related to DATASCI/STATS 531.

  • Please share knowledge about cluster computing between group members, and/or on piazza, to help everyone who wants to learn these skills.

  • Cluster-related questions can also be emailed to the U-M Information and Technology Services helpdesk, help@umich.edu

Acknowledgements and License

This course and the code involved are made available with an Creative Commons. A list of acknowledgments is available.