James and Psychohistory (was Rumble!)
deafbox at hotmail.com
Sat Sep 27 18:47:02 PDT 2003
>The best-of-breed solutions I'm aware of even try to capture social
>context, usually by data mining and extraction of social "features" from
>historical news flow.
I would expect demographic and economic
data to be more important to this kind
of modeling. And there is a LOT of
economic data to be incorporated or
used for validation, everything from
how many pallets are made last week to
how much electricity was consumed by
New York City. Some of that economic
data might serve as a good surrogate for
>The macro-economic cycle prediction stuff I'm aware of is painfully
>inadequate, on the order of accuracy of, say, prediction of weather
>specific weather at a specific location this time next year.
In some sense, weather prediction is
easier at the coarse-grained level,
because it is cyclic. Central Texas will
be hot next July, with intervening
thunderstorms. Hurricanes will strike
the southeast US sometime next fall.
People who say that we can't predict the
weather many months in advance are
overlooking the obvious.
The economy is more path dependent, and
its cycles are not nearly so regular.
And the timing of the big cycles is the
first piece of prophecy that we would
like. James sees continued growth the
next few years. I hope he is right,
because that means that unemployment
and debt don't derail the recovery. But
were those answers from models updated
with information about those events??
Help protect your PC. Get a FREE computer virus scan online from McAfee.
More information about the FoRK