Last month, I discussed the importance of variable selection as a key component of the modelling process. We examined three techniques: factor analysis, correlation analysis, and clustering. This month, we will explore CHAID/Decision Tree, Exploratory Data Analysis(EDA), and Stepwise Regression as other techniques in selecting the appropriate variables for a given model. EDA presents a more visual rather than statistical perspective of how a given input variable impacts the target variable. EDA reports depict how a given input variable is trending against the target variable. The practitioner can then select those variables which visually exhibit a trend and exclude those
To view this content
OR subscribe for free
Already receive the Predictive Analytics Times emails?
Click here to complete this one-time subscription upgrade
As of January 2014, the Predictive Analytics Times now requires legacy email subscribers to upgrade their subscription - one time only - in order to attain a password-protected login and gain complete access.