[from Jeff Vancouver 990429.0930 EST]
Gang, I am looking for the e-coli algorithms. Specifically, I am looking
at how you program reorganization.
Bill,
Regarding the modeling of instructions, I took your advice. I am not
trying to model the entire loop that determines the lower-level goal
(although this still bothers me). When I have a draft of the paper, I
would love to run it by you (indeed, I would like to make you a co-author
as I intend to use your thermos model as well). It will be at least a
month though.
Sincerely,
Jeff
[From Bill Powers (990430.1434 MDT)]
Jeff Vancouver 990429.0930 EST]
Gang, I am looking for the e-coli algorithms. Specifically, I am looking
at how you program reorganization.
See Rick's paper in Mind Readings.
Regarding the modeling of instructions, I took your advice. I am not
trying to model the entire loop that determines the lower-level goal
(although this still bothers me).
Why? Wouldn't modeling that loop just require you to account for _its_
reference signal, so you have to model the next loop up, also with no data
to enable you to check the model? I repeat, if you can determine from the
data whether the person actually did establish the reference condition you
described, you don't have to explain how that happened before you can test
the control system that receives the reference signal. If we had only the
choices of modeling the entire system in all its complexity or not modeling
it at all, we'de be stopped before we could start.
When I have a draft of the paper, I
would love to run it by you (indeed, I would like to make you a co-author
as I intend to use your thermos model as well). It will be at least a
month though.
No co-authorship, please. All you need to do to use the thermostat model is
to include a footnote, "Model courtesy WT Powers", or something simple like
that. No need to give away the store.
Best,
Bill P.