FinalProject Instruction

BUSI650
Final Project by Robin Teotia
Use of tools:
• Recommended: Excel for PART-A.
• Recommended: Python for PART-B.
Team Name:
Section:
Team Members:
Student Name ID
Forecasting
Part-A
Business Situation: ArcelorMittal Dofasco is a steel company based in Hamilton, Ontario, Canada.
Electrolyte Tinplate is one of their products. They have their manufacturing furnaces located in Calgary
and Mississauga. In order to produce this type of Tinplate they run these two furnaces once a month
for 1 hour. They have their package with the hydro of the province for 900 MWh of power per hour
which they pay yearly for both furnaces. Since their machines at both the furnaces are growing old,
they are usually utilizing above 900 MWh of Power. For each MWh unit of power above 900 units, they
must pay $7 extra.
Because of the slow in the market, they are looking to produce less Tinplate product, so they need less
power, and decided only to run one furnace either at Calgary or at Mississauga. Their estimated is
about 850 MWH.
They want to be energy efficient and asked you to do their business forecasting of power needed for
the month of September 2023 base on the given data set “Forecasting data-Dofasco”.
What would you suggest to them which furnace they should use on the basis of your forecasting? So
that they pay fewer extra charges for power.
Task1: Perform 3 month moving average, weighted moving average and exponential smoothing on
both the historical furnace power data.
Page 1 of 5
Task2: Based on your methodology and forecasting, suggest which method is good to do forecast for
each furnace.
Task3: As business analyst give your recommendation on which furnace to use for the month of
September. Support your answer with the calculation on power usage and extra cost.
Note:
• You must submit your excel file with the final pdf.
• Every time use new sheet in the workbook to perform a new method.
• Maintain the original given data as it is. It means copy it to the new sheet in the workbook and
then start working.
• You can add columns/plot/ chart… as per your requirements.
Predictive Modelling
Part-B
Task1: Select data by your choice and decide what you want to predict and with the help of what
variables.
Providing 4 data sets:

• Each group should choose any 1 out of 4.
o Boston Housing Dataset
▪ Data info/data dictionary- GitHub - eric-bunch/boston_housing
o Mtcars Dataset
▪ Data info/ data dictionary- mtcars.csv · GitHub
o Loan_data
▪ Data info/ data dictionary- easy to understand see the data.
o Real-estate1 Dataset
▪ Data info/ data dictionary-easy to understand see the data.
Task2: Perform exploratory data analysis (EDA) of data.

Task3: Define your hypothesis around modelling.
Task4: Perform simple and multiple linear regression to predict the target.
• You can define certain statistics discussed in class like null hypothesis and alternate
hypothesis for both simple and multiple linear regression.
• If in case, you are dropping any variables then on what basis.
• You can support your modelling with the help of your modelling equation.
• Etc.
Task5: Use proper metrics to evaluate your results. Define what metrics you are using, provide
explanation with formula.
Page 2 of 5
Part-C
Task6: Use any other algorithm of your choice for regression and apply it on the above data you
decided to work on.
Providing the name of algorithms:

• Ridge
• Lasso
• Random Forests
• Decision Tree
• Bagged Tress
• XG boost
Each group should choose any 1 out of the above 6 given regression algorithms.
Task7: Explain this applied algorithm in detail. (1 page max)

Task8: Explain your new results.
Important Details
Project Report:
• Must follow APA style format. (UCW Masters’ guidelines)

• Maximum 7-10 pages including everything, except appendix. Appendix can be outside the limit.
• Project Body
o Front Page
▪ Group, members, Title etc… details.
o Introduction
▪ Provide details about your project in nutshell.
▪ Tools you are using etc.
o Hypothesis
▪ What is your hypothesis behind doing this work.
• 2 hypothesis
o Around forecasting
o Around predicting
o About Data
▪ Give details about your data
o Background
▪ What algorithms and evaluation metrices you are using, provide their details.
▪ New algorithm details.
o Methodology
▪ This is basically the steps you follow to perform the task
o Results
Page 3 of 5
▪ What are your results, what features you find important, any other details…
o Discussion
▪ Explain your thoughts around the project any hurdle.
o Conclusion
▪ Generalise your results in common language, so that everyone can understand
that, even the people from non data background.
▪ What in future can be done with this data.
o References
o Appendix
▪ Provide link of the google colab here:
• Link: (make sure the link is working before submission)
▪ You can also provide some coding as well if you want. (Optional)
▪ Any other thing if you feel help explaining your project you can provide here.
(optional)
Submission Instructions:
• Pdf Report of your project.
• File nomenclature: GroupName_BUSI650_ Section_UCWID1_UCWID2…
o Eg. If the group have 5 members, then your naming should be like
▪ Numpy_BUSI650_11_33333_44444_55464_746587_758578
For Python:
• Only colab link is sufficient provided above, no need to upload any python code file.
For Excel:
• Save your excel file in the same format-
o Eg. If the group have 5 members, then your naming should be like
▪ Numpy_BUSI650_11_33333_44444_55464_746587_758578
Marking Rubrics:
Activity/Competencies Demonstrated % of
Final Grade
Critical Analysis and Research (20%)
Depth of background summary 5
Appropriateness of research 5
Task accomplishment. 10
Page 4 of 5
Statistical Explanation (40%)
Uses statistical keywords clearly and effectively 20
Information organized intelligently and to the points (avoiding 10
redundancy)
Proper presentation with explanations 10
Application of python and Excel (40%)
Coding Understanding and application 15
Excel understanding and application. 15
To the point answers (No redundancy) 10
Total 100
*Every question has equal weightage, not attempting all the questions will impact marking rubrics
proportionally.
Remark: For any question/doubt, feel free to ask on discussion forum for final project.
Page 5 of 5

FinalProject Instruction

Uploaded by

Copyright:

Available Formats

FinalProject Instruction

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

FinalProject Instruction

Uploaded by

Copyright:

Available Formats

BUSI650

Final Project by Robin Teotia

Providing 4 data sets:

Task2: Perform exploratory data analysis (EDA) of data.

Providing the name of algorithms:

Task7: Explain this applied algorithm in detail. (1 page max)

• Must follow APA style format. (UCW Masters’ guidelines)

You might also like