From Greg Williams (921221 - 2)
Bill Powers (921221.1500)
The PCT model is fundamentally a generative model. As such it is
only partly successful. It will become more successful as we
become able to simulate more and more complex behaviors, thus
showing that the structure of the model is plausible.
This is the direction in which I was trying to point when I asked earlier
today about modeling step tracking performance. Given good prediction of
tracking with low error over most of the duration of a trial, I think the
logical next step is to attempt prediction when error is higher over most of
the duration of a trial. Also, because good control obscures system
parameters, I think higher error situations will make it easier to distinguish
among candidate models with regard to their predictive abilities, and to
identify nonlinearities in system components.
By using a little control
system, the program adjusts the difficulty of a task (by varying
the speed with which a table of disturbance values is scanned)
until a specific amount of RMS tracking error is produced by the
participant. This amount of error is then maintained quite well
in a subsequent one-minute tracking task. The purpose is to
measure parameters of control at standard levels of tracking
error, and also to monitor long-term changes in tracking skill.
Sounds great -- in some ways better than steps/ramps, but I suspect figuring
out which types of nonlinearities to try in the models might be easier with
steps/ramps. Also, be sure to look at a broad range of disturbance bandwidths,
to make sure the "good" models aren't biased for a notch which you are
gradually moving up the spectrum. I'll be interested to see your results.
What's the timetable?
Should be easy for somebody who can predict for One whole
minute, right?
Yeah. When are you going to do it?
I don't need to, since you already are.
As ever,
Greg