Case 6

As a team of 2 to 3 students, develop your own case study/big data analysis based on real data. The perfect case would have the following attributes:

  • Large set of real datathat is relevant to you somehow(not just what you could find on the internet). Maybe from work, sports, hobby, etc. As a last resort, I have some data sets that I can provide you with.
  • Uses multiple techniques that we have learned in class:
  • Modeling a probability from data.
  • Curve-fitting.
  • Non-linear Forecasting with error terms.
  • Optimization using Excel solver
  • Deals with outliers.
  • Something like the zip codes where your data is mapped or grouped by location.
  • Data conversion, where you may have to convert or clean messy data.
  • Something like the Zillow.com case where you try to out-predict a professional website.
  • Ambitious: A very difficult, real problem with only a partial solution will be considered a greater success than an easy problem with a complete solution. (By the way if you just put a straight line through 10 data points you will not only fail this assignment but the class, too.)
  • The right amount of informative graphs and charts.
  • Touches on all aspects of Business Analytics: descriptive (graphs), predictive (forecasts), and prescriptive (decisions/answers).
  • You will be proud of your final results. The final paper you write up may not be publishable, but it should be close.
  • Depending on time, you may be required to give a brief (3 minutes, 3 slides) presentation of your results to the class

Part 1

Due by the end of class today:

Brainstorm ideas with your partners. Submit a typed memo with the names of the team members, the main idea you want to pursue, where you will get your data and what kind of analyses you plan to perform, which analytical techniques you plan to use and any questions you have. Also list a couple of backup ideas in case your main idea doesn’t pan out.

Go big! If you pick a topic you’re interested in, you may come up with something really cool and might even enjoy doing it. [On the other hand, if you pick a topic that is dull to begin with, it’s not going to get more interesting as you start crunching numbers on it.]