overfitting

(As for Jeffrey's 1968/2008 presidential campaign parallel ... it's not a bad exercise, but in the main I think Jeffrey's "overfitting" the data too muchly.)

Archive 2007-11-01 2007

(As for Jeffrey's 1968/2008 presidential campaign parallel ... it's not a bad exercise, but in the main I think Jeffrey's "overfitting" the data too muchly.)

Your Right Hand Thief 2007

This is completely consistent with our understanding of MBH methods – a classic spurious regression of the bristlecones combined with massive "overfitting" from essentially white noise series in the calibration period, thus the breakdown of the cross-validation R2.

Rutherford et al [2005] Collation Errors « Climate Audit 2006

The statistical issue that it turns over – and it’s something that isn’t considered anywhere in multivariate statistical literature – is what happens when you mix one “classical” univariate spurious regression with multivariate overfitting which is my model for MBH.

Plant Deposits at Quelccaya « Climate Audit 2006

As in Brown’s multivariate calibration confidence formula, which takes this kind of overfitting large amount of responses into account in the uncertainties.

Overfitting by Inverse Regression « Climate Audit 2006

Avoiding that disaster — called overfitting — is a bit of an art; and being very good at it separates masters like BellKor from the rest of the field.

February « 2008 « Isegoria 2008

Avoiding that disaster — called overfitting — is a bit of an art; and being very good at it separates masters like BellKor from the rest of the field.

This Psychologist Might Outsmart the Math Brains Competing for the Netflix Prize « Isegoria 2008

Then again you can do much worse than just using normal assumptions (maybe autochecking a few heavy-tailed/one-sided/clever alternatives using some information criterion against overfitting) and churning on that. 95% of the result in 5% of the effort …

NeuroBayes: Sometimes sans Neuro « Imaginary Potential 2008

It's nonsense; any QSAR modeler can explain in details the dangers of overfitting, extrapolation, etc, etc.

Archive 2008-08-01 Egon Willighagen 2008

It's nonsense; any QSAR modeler can explain in details the dangers of overfitting, extrapolation, etc, etc.

"The End of Theory: The Data Deluge Makes the Scientific Method Obsolete" Egon Willighagen 2008