Appendix A: Phase 1 - Business Understanding -

oClearly state the Business objectives and data mining goals.

oClearly state the business question(s) that you are trying to address

oClearly state the data mining techniques you are performing to address the business question(s).

Appendix B: Phase 2 - Data Understanding

oDescribe data - distributions (normal or skewed) of the continuous variables, discussion of missing data, ...

oExplore data - Graphs that will help you understand the relationship between the variables in the data sets. Should help you provide information and insight. Explain the graphs accurately. Develop eight graphs - some simple & some complex. Use "Graph Builder" tool to generate graphs. Do not run any statistical analysis here.

Appendix C: Phase 3 - Data Preparation

oCleanse data if necessary and provide the details - recoding of data, removal of duplicate data, exclusion of extreme outliers, imputing missing data, etc.

oFormat or transform data if necessary, create new variables if necessary, etc.

oPlease explain clearly what you did and why you did it.

Appendix D - Phase 4 (Model Building)

oPartition the data - use 75% data for Training, 25% for Validation - use Stratified Random on Churn

oBuild the two models using "Churn" as your dependent variable and explain all the results from the JMP run in detail, with appropriate screens copied from JMP run. Interpret the results and provide your conclusion.

oNotice that this is the most important phase of your paper. Provide detailed analyses and try to address the business questions based on the results.

Appendix E -Phase 5 (Model Validation)

oDiscuss how you validated different models and how you picked the final model in this section

oCreate confusion matrices for different probability cutoffs (for both LR and DT)

oCompute the models for each scenario

oSelect the best LR and the best DT models and clearly outline how you selected the best model.

oSelect the best of the best model for implementation

Appendix F -Phase 6 (Implementation)

oIn this section provide your conclusion and how the solution or model should be implemented.

oIdentify the profiles of the customers who are likely to Churn.

oExplain how to you would use the results of your model in developing a new marketing plan that can reduce Churn.

Integrate JMP output at the appropriate places in your report where you discuss the results. Your report should be professionally looking. Proper use of heading, formatting, and structure is very important.