###   
###
###  Spring 2020 - Final Project
###
###
####################################


nmimoto@uakron.edu


1. Form a gorup of up to 4 students.  



2. Pick a dataset to work on.

  You can use dataset listed under "Data List" on class webpage,
  or you can find something on the web. If you happen to have
  time series data (say, from your work) you can use them as well.

  The dataset must be approved for the project. When you make a decision, email me.




3.  You are exptected to do:

  Find best SARIMA model that fits the data.

  Write report in MS Word explaining how you arrived at the final model.

  Report can not be more than 5 pages, including R outputs, and figures.

  Key output, and key plots should be included, but you must carefully
  choose what to include. Don't just include all the plots, or just
  omit all the outputs.  

  All R code should be attached at the end.  R code deos not count toward 5 pages.




4.  You can refer to Modeling Guidelines.

  Below are list of general questions that you need to answer in your report.  

    Where did you get the data? 
    Seasonality?
    Stationality?  Do you need d=1? D=1? Both?
    Need Transformation? (log?)
    Linear Trend or Random Trend?
    Order selection of ARMA(p,q)x(P,Q)
    Residual Analysis

    (if appropreate)
      Sign of over-differencing?
      Sign of under-differencing?
      Is one model better than the other? why?




5. You will be graded for:

  Clarity of explanation of how analysis was conducted.

  Whether you used appropreate statistical method for given situation.
    
  Correctness of your interpretations of the outputs from statistical tools.

  You will not be graded on whether your final model is actually
  the best model for the dataset or not.  




6. I am available for consulting via email and WebEx meeting.

  You can also submit draft/partial report to me to get feedback.
  (I need at least two days to return the feedback.)



7.  Report must be submitted in Brightspace.

  If you used dataset that is outside of R package or the class website, then
  you must submit the dataset as well.  



8 DUE DATE:  May 1st 11pm 

  You are strongly encouraged to submit the report by above due, but
  late submissions are accepted until May 8th 11pm.  



Any questions, email nmimoto@uakron.edu.