Glen E. P. Ropella wrote:
> Another example is an RC
> plane versus a balsa wood plane as models of a life size plane.
> All that's required is a common aspect ("lift").
Yes, I agree that it's better to have many models of something than just
one, as that will tend to explore at least some `interstitial space'.
No, I don't agree that just because there are multiple models that those
models won't be prone to correlation. People are copycats, in subtle
and not so subtle ways. And as Eric points out, even a pure
statistical inference approach is prone to correlation that will require
further investigation to get at causality. To my way of thinking, data
mining is just a first step to get into the right space at all. My
concern is simply the one that Phil raises from time to time: Looking
for your keys under the street light, when in fact they are in a dark
alley.
============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
lectures, archives, unsubscribe, maps at http://www.friam.org
