Friday, July 17, 2009

Inductive Bias is Redundant

This exchange, on PSJR, caused me to shoot coffee out my nose.

Dork 1. Apologies to the quants if this is a naive question, but as someone just coming to quantitative methods I was wondering if there is anyone currently using machine learning algorithms to study political data?

Yes, I realize machine learning is going to be over my head by several decades, but I am excited about anything that promises to help manage inductive bias.

Dork 2. Define "inductive bias". How is it different from "bias"?

Comic Genius Quant Type 1.
Define "inductive bias". How is it different from "inductive"?

1 comment:

SB7 said...

Ha!

I don't know about you PoliSci types, but us machine learning people actually have a specific technical definition for "bias." Perhaps that is confusing one or both of the Dorks?