This post is part of the series, Deep Dive.
As we know, building a model in Predict is incredibly easy. Interpreting the results and confirming that the model is accurately predicting the outcome is a whole other thing. In this on-demand video we dive into the actual model diagnostics and discuss not only what these different validation statistics are, but how we can use them to help identify goodness of fit. We’ll look at the Decile Analysis, the Wald Chi Squares, the Concordance rates, and other values to see how they can be used in checking model accuracy.
Keywords: Predictive Modeling, Decile Analysis, Wald Chi Square, Percent Concordant, ROC Curve, Lift, Model Accuracy
Want to follow along? Download the .VPA analysis files here!