Forecasting: Forecasting impacts so many parts of business. It is used to set budgets, allocated resources, plan for the future, etc. There are many ways to forecast since there so many nuances to it. The forecasting we will look at this semester is based on the historical momentum of the data being forecasted, using various techniques like period-over-period growth, regression, etc. The assignment here is to apply the best technique to forecast a real business dataset.
The project requires you forecast the historical data, replicating the appropriate trends, seasonalities, week-over-week variability, etc. Please refer to the chapter and materials on Forecasting for the procedures to do.
The Questions
For each question, show the work to support your answer.
Analysis
a) Using regression as the procedure, produce the forecast of weekly sales for the periods Jan 27, 2014 through Jan 19, 2015
b) Show a graph of the historical and forecasted sales in one graph
c) Which week in the forecasted period is expected to have the highest volume of sales and which week is expected to have the lowest volume of sales
d) Using year-over-year change as the procedure, produce the forecast of weekly sales for the periods Jan 27, 2014 through Jan 19, 2015
e) Show a graph of the historical and forecasted sales in one graph
f) Which weeks in the forecasted period are expected to have the highest volume of sales and the lowest volume of sales (5 points)
General
e) Does either of the forecasts "look" good to you? Why or why not?
f) What are the implications to a business if the forecast is too high?
g) What are the implications to a business if the forecast is too low?
h) Give three examples of of data that cannot be forecasted and why.
Download:- Forecasting_project.xlsx