Binary Classification Class Project Flow Overview
Binary Classification Project Flow Step 1 Variable Selection Primary Issue: Little knowledge of content Step 2 Variable Cleansing Primary Issue: Mixed Variables Step 3 Variable Reduction Primary Issue: Multicollinearity STAT4330/8330 PRIESTLEY
Binary Classification Project Flow Step 4 Discretization Process One (equal width categories) Primary Issue: Determining user defined cut points for ordinal values Step 5 Discretization Process Two (equal frequency categories) Primary Issue: Understanding the output from SAS Step 6 Model Development Primary Issue: Interpretation of beta coefficients STAT4330/8330 PRIESTLEY
Binary Classification Project Flow Step 7 Model Evaluation Primary Issue: Interpretation of ROC, Profit Function, Gains Chart, Classification Table Step 8 Behavior Analysis/Cluster Analysis Primary Issue: Method to determine behavior of good/bad Step 9 Report and Presentation STAT4330/8330 PRIESTLEY