This course is for junior/senior undergraduate students. The main objective of this course is to provide you with the proper foundation to analyze and forecast time series data in a professional setting. This means that in addition to forecasting into the future and evaluating such forecasts we will discuss other topics to prepare you for your journey as an analyst. A survey of analytical techniques used in forecasting. Techniques include exponential smoothing, ARIMA, multiple regression, and judgement-based methods. Implementation issues and challenges are discussed.
- Name: Zhaohu (Jonathan) Fan
- Title: PhD Candidate, Department of Operations, Business Analytics and Information Systems
- Office Information: LCB, Room 3327
- Email: [email protected]
- Office Hours: Monday & Wednesday 11:00 AM to 12:00 PM and by appointment
Communication Policy: Students are encouraged to contact me anytime via email or phone. Please use email as the primary mode of contact. A response will be given within 36-48 hours. Please understand that I cannot guarantee an immediate response if you contact me very close to an assignment deadline.
- Provide students with a foundational knowledge of time series analysis
- Expose students to a number of traditional and contemporary methods in time series and forecasting
- Familiarize students with many of the challenges associated with time series forecasting
- Provide students with practical experience analyzing real-world data and communicating the results
While I will try to focus on the application over the theory to maximize the above objectives, I will provide additional optional reading for those interested in a deeper dive into the theory 🚀.
- have a solid foundation for approaching time series analysis and forecasting projects in the future
- provide of project for portfolio
- practice with R programming language as it relates to time series data
Module | Description |
---|---|
1 | Module 1 – Practical Time Series Analysis (weeks 1-3) |
Welcome Lecture and Notes | • Why is time series analysis and forecasting important? |
Practical Time Series Analysis | • How to approach a forecasting problem |
R lab-I | • Visualization and exploration of time series data |
R lab-II | • Wrangling with time series objects |
R lab-III | • Features of time series data |
2 | Module 2 – Forecasting Basics (weeks 4-6) |
Forecasting Basics-Part I | •Four simple forecasting models |
Forecasting Basics-Part II | • Forecasting process |
R lab-IV | • R lab (Basic tools For forecasting-I) |
R lab-VI | • R lab (Basic tools For forecasting-II) |
R lab-VII | • R lab (Evaluation of model performance-I) |
3 | Module 3 – Forecasting Models (weeks 7-10) |
ARIMA-Part I | • ARIMA models-I |
ARIMA-Part II | • ARIMA models-II |
ARIMA-Part III | • ARIMA models-III |
ARIMA-Part IV | • Seasonal ARIMA models |
R lab-VIII | • R lab (Evaluation of model performance-II) |
R lab-IX | • R lab (ARIMA-I) |
R lab-X | • R lab (ARIMA-II) |
4 | Module 4 – Additional Topics (weeks 11-14) |
Additional Topics | • Exponential smoothing-I |
R lab-XI | • R lab (Simple Exponential Smoothing) |
R lab-XII | • R lab (Holt's Method) |
R lab-XIII | • R lab (Holt-Winter Method) |
-
Exams – There will be two exams given during class after we finish module 2 and module 4 to assess grasp of key concepts of time series analysis and forecasting.
-
Assignments – 4 take-home assignments will be given throughout the semester. Students will have 2 weeks to complete the assignments.
-
Discussions – The first 14 classes will have in-class labs/discussions. Write-ups from these discussions/labs are to be submitted on Canvas by 11:59pm on the Sunday after class. If you cannot make the class, these will be posted and submission will be allowed.
-
Project – The project is a core component of this course. Students will pick a topic and dataset of their choosing to analyze and forecast with the methods taught in this course (and ideally some additional methods). These topics will be submitted during week 6 and the analyses will be presented during week 15. In lieu of a final exam, students will submit a reproducible markdown/notebook.
Class video, homework and class projects are available in Canvas. Please check homework in Canvas and submit all your homework through Canvas. All announcements are in Canvas.
Acknowledgments: I have drawn ideas or readings from the following texts:
- Rob J Hyndman and George Athanasopoulos, Forecasting: Principles and Practice (3rd ed)
- Bradley Boehmke, UC BANA 6043 Statistical Computing
- Ethan Swan, Python for Data Science
- And many more.