Econ 345: Applied Econometrics Forecasting Competition

Econ 345: Applied Econometrics Forecasting Competition
The forecasting competition is designed to give you real-world experience of forecasting. As part of the
course we are having a real-world forecasting competition – using a model you estimate, you will be
predicting time series observations. However, forecast accuracy will not affect your grade. As part of a
group, you will submit your forecasts online, together with a short essay (maximum 2 pages) describing
your forecasting model and resulting forecasts.
The task is to predict the number of cyclists on the Galloping Goose trail from Friday Nov 26
th to
Saturday Dec 4
th (inclusive, i.e. 9 days). The forecasting competition consists of two parts. Both parts
have to be completed to obtain full marks:
1. Write a short report describing your model and resulting forecasts.
2. Submit your forecasts of the number of cyclists over via Brightspace as a .csv file by November
25
th
, 4pm.
You may work in groups and write one report (and produce one set of forecasts) as a group, but each
student must submit their own copy of the report on Brightspace. Please make sure to include the names
and V-numbers of all group members on the report.
Details on both the report and submission of the actual forecasts are provided on the following two
pages.
Data to be used as the basis for your forecasts will be posted in, November, however, you can already
use the Galloping Goose data on Brightspace to estimate models using last year’s data.
1. Details on the report to be submitted:
As a group, prepare a short report on your forecasting model and resulting forecasts. This should be
submitted alongside your forecasts via Brightspace. You may work in groups and write one report as a
group, but each student has to submit their own copy of the report on Brightspace. Please make sure to
include the names and V-numbers of all group members on the report.
The report should take the form of a short academic essay (avoid casual language, copying R-code, or
showing screenshots). A good report would:
i. Have a clear structure, such as:
a. Introduction: introduce the forecasting problem.
b. Data: describe the dataset, any additional variables you may be using, any apparent
patterns, trends etc. including any relevant references.
c. Methods: describe your forecasting model, ideally including an equation that shows
your model. Carefully specify what variables are included and why you include them.
d. Results: if applicable show your in-sample estimated model as a table (similar to the
research project formatting) and plot your forecasts in a well-formatted figure,
labelling axes, clearly differentiating between in-sample observations and forecasts.
For example:
e. Conclusion: summarise your results, discuss how your forecasting model could be
improved.
ii. Be well-written: do not use casual language, this should be an academic piece of writing.
iii. Be well formatted and presented: do not include screenshots of R-output, do not include
R-code. If you present regression results these should either be in the form of an equation
(as we have seen in lecture slides) or as a table, and include coefficient estimates,
standard errors, and the number of observations.
iv. Be clearly referenced. Make sure to include references to any relevant datasets or
literature.
v. Not exceed 2 pages in length (excluding references).
Figure 1: Observed (black) and predicted (red) cyclists on the Galloping Goose.
2. Details on the Forecasts to be submitted:
Upload your forecasts as a “.csv” file to Brightspace (under `Entry for Forecasting Competition’). The
forecasts must be in a csv file and must take the form shown below – the first column has to be the date
using the format provided below (labelled as date), the second column the number of predicted cyclists
(labelled cyclists) from 2019-11-26 to 2019-11-29, and the third column must contain all V-numbers
of you and your team members. The fourth column should be labelled “team_name” and in the first row
contain your chosen team name (you are free to choose any name you want).
Only one copy of the forecasts has to be submitted per group – not every team member has to upload
the groups forecasts to Brightspace.
Forecast accuracy will not affect your assignment grade. Forecast accuracy will be judged by the lowest
RMSE over November 26
th
-November 29
th. Good luck!
Format of csv file:
date cyclists v_numbers team_name
2019-11-26 254 (your prediction here) V00001 Myteam
2019-11-27 256 V00002
2019-11-28 …. ….
2019-11-29