Keith McCormick

A former training participant emailed me recently.  
They asked the following: "I have come across an issue that I would like to ask you about; the issue concerns C5.0 and multicollinearity.  I am completing a project in which I used the C5.0 method of modelling.  We checked for multicollinearity early on and took out the variables that were […]

CRISP-DM is nearly 10 years old.  As a result it needs some attention.  I just returned from the SIG meeting in London last week.  The small, but spirited group had lots of advice for the consortium.  The main thing to report is that 2.0 is targeted for summer.  Also, the group seemed to think that […]