In my last two posts I described why overfitting predictive models is dangerous beyond the most obvious problem, namely that accuracy on new data is lower than expected. In the next few posts, I’ll describe how to recognized that overfitting may be occurring, and some common approaches to remove or mitigate the effects of overfitting. OVERVIEW Overfitting is perhaps the most common and destructive problem in predictive modeling. It is common because predictive modeling is often an inductive, data-driven exercise where the data is king, as opposed to threads of statistical modeling where the model is king (terms borrowed from
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.