STAT 496 Applied Time Series Analysis Project
Due by Dec. 12 (or earlier)
This project can be individual or in groups of up to 3. The groups do not need to be the same as your HW groups.
Having trouble finding a dataset? Here are some useful sources
Objective:
The objective of the project is to demonstrate and present time series data analysis
concepts from the class (or literature) to a problem or question of interest in
a dataset of your choice.
The project involves data and a problem.
The problem will (usually) involve: 1) fitting an appropriate time series model and 2) forecasting (some reasonable
step-ahead forecasts).
Guidelines for 5 minute presentation: (on M or W, Dec. 10 or 12)
You are allowed 1 slide. (You can use the white board, if necessary!)
- Description of your data.
- Description of the problem of interest.
- Description and justification of model(s) that you have selected.
- Results of the best model.
Guidelines for Written Report: due on W Dec. 12
Your report should include:
- Abstract or Summary
- Introduction
- Methodology
- Results
- Conclusions, etc.
- Bibliolography (if needed)
The report should be 3-5 pages, 12 point font, and
single-spaced. DONE IN SWEAVE!
The 3-5 pages does not include figures, tables, nor bibliography.
More details of the Sections of the written report are as follows:
(There are also more complete guidelines linked at the bottom if you would like more information)
- Abstract or Summary. This section should be a one paragraph summary of the report.
- Introduction. This section should provide a background for the problem.
It should contain a description and objective of the problem. Are there any
expectations based on literature or scientific knowledge?
Remember to explain the problem you are trying to solve!
- Methodology. Explain how the data was collected. List the variable(s) and
give a description of each variable. You can include a time series plot of the data. Explain what methods you used in your
analysis. You can use your textbook Section 1.2 and follow the three main steps.
(This is actually called the Box-Jenkins Method). Then you can describe the forecasting that you perform.
- Results. Here is where you present the results of the best model.
Give the results
of your statistical analysis (hypothesis tests, p-values, estimates, CI's, ect.).
Your R code and R summaries should be in the Appendix.
Include the final model diagnostics plots. Include the plot of the forescasts with forecast intervals.
(If you made a lot of plots in order to specify your model, you can include them in the Appendix.)
- Conclusions, etc. This is your discussion and interpretation of findings.
- Appendix. Your report should be understood independent of the
Appendix. For this project I would like you to include a print out of the
beginning of your dataset, so that I can see the structure of your dataset.
Include your R code, summaries, and plots used in your analysis.
Here is link to Guidelines for the Written Report giv\
es details.
(Here is link to Professor Kafadar's complete guidelines)