Glen E. P. Ropella wrote:
> I don't imply that approximations cannot be obtained by taking various
> slices of X {x1, ..., xn} and Y {y1, ..., ym} and examining the
> sub-inference from xi -> yj. But, there will always be room for
> skepticism that your particular slices adequately capture the cause and
> effect relationship.
A model either gives an edge on prediction or it doesn't. It is
quantifiable provided there is consensus on the available variety of
available input and output measurements and many such measurements. If
someone wants to be skeptical of a model that has never made a bad call
in a millions of circumstances, that's fine but at some point one has to
wonder if they are philosophy students.
============================================================
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
