Stat 6550: The Statistical Analysis of Time Series (Spring 2016)

Peter F. Craigmile

Email: pfc <at> stat.osu.edu
Office: Cockins Hall, Room 427

This class aims to develop a working knowledge of time series analysis and forecasting methods. The emphasis is on modeling methodology (identification, estimation, diagnostics, and updating) and forecasting. Experience is gained in the statistical theory so as to be able to analyze time series data in practice.

Prerequisites: Stat 6201 or Stat 6302 or Stat 6802; Stat 6450 or Stat 6950; or permission of instructor.

Notes

All class notes will appear here as PDF files.

  1. Analyzing time series
       (2 per page) (R code)  (posted Tues 5 Jan) 
  2. Time series models and stationary processes
       (2 per page) (R code)  (posted Tues 12 Jan) 
  3. Estimating the mean, ACVF and ACF; Q-Q plots
       (2 per page)  (posted Tues 19 Jan) 
  4. Handling trend and seasonality
       (2 per page)  (posted Tue 26 Jan) 
  5. Stationary and Linear Processes
       (2 per page)  (posted Thu 4 Feb) 
  6. Predicting and forecasting stationary time series
       (2 per page)  (posted Thu 11 Feb) 
  7. Defining ARMA processes
       (2 per page)  (posted Tue 23 Feb) 
        Handout: Calculating the theoretical ACF and PACF in R
  8. Modeling ARMA processes
       (2 per page) (R code)  (posted Tue 8 Mar) 
  9. A modeling case study
       (2 per page) (R code)  (posted Thu 24 Mar) 
  10. Modeling ARIMA and SARIMA processes
       (2 per page) (R code)  (posted Mon 4 Apr) 
  11. Regression with time series errors
       (2 per page) (R code)  (posted Tue 12 Apr) 
  12. Some interesting topics in time series analysis
       (2 per page)  (posted Tue 12 Apr) 

Appendices

  1. Useful math results
       (2 per page)  (posted Mon 4 Jan) 



[Return to the teaching page]