Posted by
glen e. p. ropella-2 on
Sep 16, 2009; 9:41pm
URL: http://friam.383.s1.nabble.com/Re-comm-was-Re-FW-Re-Emergence-Seminar-BritishEmergence-tp3654051p3659301.html
Thus spake Miles Parker circa 09-09-16 12:49 PM:
> I would heartily agree -- and as the name of this group is "Applied
> Complexity" -- that that is sound practical advice. Curious if you are
> also arguing that *in general*, say
>
> validity(M1) > validity(M2)
>
> where M1 and M2 are sets of models and M1@ > M2@, or M1 ⊃ M2?
I'm not sure what that last symbol is (⊃); but I do NOT assert that
any set of models that's larger than any other set of models is more
valid merely because it's larger. Rather, I assert that more models are
required to represent a complex system than a simple system. And adding
another (valid) model to a set of (valid) models of a system increases
the accuracy of the set.
It's important that each model be validated against some aspect (use
case) of the system, of course. Adding an invalid model would probably
-- but not necessarily -- decrease the validity of the whole set. Also,
it's necessary for the models to be validated against different aspects
of the system. I.e. just adding another model that validates against an
identical aspect as another model already in the set probably won't
increase the validity of the whole set (by much anyway).
> I think but am not sure that I am interpreting your parallax argument
> correctly. Do you mean that literally, or as a kind of heuristic? Does
> it assume a known or knowable state-space?
I mean it literally. It does not assume a known or knowable
_mechanism_. It does assume known or knowable phenomena (which is a
fancy way of saying that we have to have concrete _measurement_
operators we apply to the model and the system for validation).
Much of this is spelled out in excruciating detail in our newest paper:
http://www.springerlink.com/content/v6q24gw11l037351/Sorry for the plug.... I didn't intend for the conversation to go this way.
--
glen e. p. ropella, 971-222-9095,
http://agent-based-modeling.com============================================================
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