Re: No pure predictors (was: Singularity Blues)

From: Simon Belak (
Date: Sun Apr 10 2005 - 17:06:05 MDT

I need to amend myself somewhat. In fact a decision making mechanism is needed
in cases when a number identically fit options are presented.
Two functionalities have to be implemented. If there exist a simulation to which
we haven't yet subjected our input data, pass it on. Prefered order is either
an emergent property of implementation or subject of another simulation or even
arbitrary. One needs to bare in mind that our main goal is fulfilled with any
of the options presented.
Else choose one upon an arbitrary (one may even say banal) criteria, for
example: "always choose the blue one".


This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:51 MDT