Explain what investigative questions you will be addressing and how these relate to possible policy recommendations. Explain why certain methods for analysis have been selected. You should use at least two methods to develop results. Explain the model selection rationale and the comparative advantage & potential weakness for each of the chosen modeling techniques.

Max 10 pages paper. Illustrations do not count as pages with the page limit.

Report Content.

Explain what investigative questions you will be addressing and how these relate to possible policy recommendations.
Explain why certain methods for analysis have been selected. You should use at least two methods to develop results. Explain the model selection rationale and the comparative advantage & potential weakness for each of the chosen modeling techniques.
Specify the dataset source and its description. Develop a list of variables actually used in your modeling. Briefly describe any new data items derived from client-supplied data and any external data that you added to the source data provided. What if any limitations existed from the source data and potential implications for the quality of the results .
Clean/transform/merge/select the data for analysis. What was done–why? Submitcleaned data input files .
Perform the DM analyses including models and additional evaluation of model output, e.g., web, gains, matrix as appropriate. Explain your results and any sensitivity analyses performed. At least 2 different models should be used in your project. Submit the final stream(s).
Discuss what you learned to confirm or disconfirm expectations any meaning ful patterns discovered from predictive analytics and/or EDA methods . What 3 recommendations are proposed to your client?
Evaluation: 1. Report 350 pts  2. Presentation 50 pts. Content will be 80% and Style will be 20%

Last Completed Projects