To Fit or Not to Fit Data to a Model
What if Shakespeare was a data scientist? Today's big data necessitates  Let the data define the model.
By Bruce Ratner, GenIQ.net
To fit or not to fit data to a model  that is the question:
Whether 'tis nobler in the mind to suffer
The slings and arrows of outrageously using
The statistical regression paradigm of
Fitting data to a prespecified(!) model, conceived and tested
Within the smalldata setting of the day, 206 years ago,
Or to take arms against a sea of troubles
And, by opposing, move aside fitting data to a model.
Today's big data necessitates  Let the data define the model.
Fitting big data to a prespecified smallframed model
Produces a skewed model with
Doubt interpretability and questionable results.
When we have shuffled off the expected coil,
There's the respect of the GenIQ Model,
A machinelearning alternative regression model
To the statistical regression model.
GenIQ is an assumptionfree, freeform model that
Maximizes the cum lift statistic, equally, the decile table.
This was originally published at http://www.geniq.net/res/ShakespearianModelogue.html.
Bruce Ratner, Ph.D., The Significant Statistician™, is President and Founder of DM STAT1 Consulting, and the author of the bestselling book Statistical and MachineLearning Data Mining: Techniques for Better Predictive Modeling and Analysis of Big Data.
Top Stories Past 30 Days  


