Scribd
Upload
138 views2 pages

MAT2440: Time Series Forecasting Assignment

This document outlines the requirements for a two-part time series forecasting assignment. Part 1 involves analyzing provided time series data using smoothing, decomposition, autocorrelation, and forecasting models to predict the next 6 values. Part 2(a) repeats the analysis using SARIMA models. Part 2(b) requires students to find their own time series data, undertake identification, estimation, diagnostic checking on competing models, and forecast holding out recent values for comparison. The final report must include an executive summary, introduction, body, conclusions, and appendix with supporting materials.

Uploaded by

Oli Guidetti
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
138 views2 pages

MAT2440: Time Series Forecasting Assignment

This document outlines the requirements for a two-part time series forecasting assignment. Part 1 involves analyzing provided time series data using smoothing, decomposition, autocorrelation, and forecasting models to predict the next 6 values. Part 2(a) repeats the analysis using SARIMA models. Part 2(b) requires students to find their own time series data, undertake identification, estimation, diagnostic checking on competing models, and forecast holding out recent values for comparison. The final report must include an executive summary, introduction, body, conclusions, and appendix with supporting materials.

Uploaded by

Oli Guidetti
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

School of Engineering

___________________________________________________________________________________

MAT2440: Time Series Forecasting


Assignment
There are two parts to this assignment.
Part 1 (50%) Due 13:30pm 7 May 2015
Undertake a time series analysis for each of the attached time series data sets using
smoothing and/or classical decomposition.
Your analysis should include:

Appropriate time series plots;


Autocorrelation Function and its interpretation.
An identification of appropriate models to fit the data.
Forecasts for the next 6 values including where possible prediction intervals
for the forecasts.

Your assignment should outline the process you used in obtaining your models
including reasons for choosing your model. A copy of computer log can be included
including analysis of models you finally rejected.
NOTE: Remember there is no one correct answer.
In marking the assignment I am particularly interested in assessing the modelling
process you undertook to obtain an acceptable answer.
Part 2 (50%) Due 9am 21st May 2015
Part (a)
Repeat the above analysis for the data sets given using an appropriate SARIMA
model.

Part (b): Applied Time Series Project.


The Data
You should each find your own time series data for this project. It is hoped that the
data will be from an area of interest to you. Ideally, the series will be of length 100200 observations. In the social sciences we may have to get by with a shorter series.
The more data available the more complex models you may consider (as long as the
model does not change over time). The source of the data should be given and a
listing of the actual values provided in the appendix to the final report and on disk.
Please provide the data in either Excel or "text" (ASCII) format. A layperson's
description of the data, what is being measured, how it is measured, etc. must be
included in the final report.
If you have trouble gaining access to your own data then you can obtain an
appropriate data set from the web site:
http://www-personal.buseco.monash.edu.au/~hyndman/TSDL/

The Analysis
Model identification should include plotting, differencing (if applicable), sample ACF
and sample PACF. Transformations such as logarithms should be considered (if
applicable). Estimation and diagnostic checking must be carried out. Competing
models should be compared.

Forecasting
Several of the most recent series values should be "held out" of the model-building
phase so that forecasts may be compared to actual values.

The Final Report


The final report should be as complete and readable as possible. The report should
have a cover page, a one page "executive summary", introduction, body, conclusions,
and appendices. All supporting graphs or computer "printouts" should appear as
exhibits and be referred to as appropriate in the narrative. Nearly all of the supporting
printouts and graphs should appear only in the appendix.
This part of the assignment will allow you to undertake an analysis of data of interest
to you or from your work. You should prepare the report in a form that could be
presented to senior management.

You might also like