Overview
Your data is a de-identified Excel spreadsheet. Your assignment is to perform the appropriate analysis to forecast several series for 140 periods. You will have 1622 periods for your analysis. See Requirement #2 for more details.
Requirement #1
You will turn in a written report. You need to write this report as if it the report will be routed in an office to personnel of vary different backgrounds. You need to be able to reach readers that have no data science background to full fledge data scientists. So, you need to explain what you have done, why and how you did it with both layman and technical terminology. Do not simply write this with solely me in mind. Visuals and output are expected, but do not to include every bit of analysis you perform. A terse report with simple terminology get a higher score versus throwing in everything into a long, ad nauseam report. This is an art and a central theme of data science – the balance to automation and math traded with simplicity and transparency. Story telling is really taking on for data science, so please flex your muscles here. The report is part 1 of 2 requirements.
NOTE: We have covered a lot of material. I do not want you to mention every method in the report. You should be sure to cover fundamentals, visualize your data, and perform exploratory data analysis as appropriate and then try a variety of appropriate predictive methods.
Requirement #2
Your second requirement is to produce forecasts. Your returned (what you hand in) workbook will contain at least 6 sheets where I will calculate your error rates. There will be one sheet (tab) for each Group – S01, S02, S03, SO4, S05, S06. You should order each sheet by the variable SeriesIND (low to high). Your source data is sorted this way, except there are all 6 groups present in one sheet which you must break out into 6 tabs. You will submit the data I sent AND the forward forecast for 140 periods. I want you to forecast the following
S01 – Forecast Var01, Var02
S02 – Forecast Var02, Var03
S03 – Forecast Var05, Var07
S04 – Forecast Var01, Var02
S05 – Forecast Var02, Var03
S06 – Forecast Var05, Var07
