Dr. P.P. Mujumdar (pradeep@CIVIL.IISC.ERNET.IN)
Mon, 21 Aug 1995 19:32:12 +0500
Message-ID: <Pine.3.89.9508211924.A3750-0100000@civil.iisc.ernet.in>
Date: Mon, 21 Aug 1995 19:32:12 +0500
From: "Dr. P.P. Mujumdar" <pradeep@CIVIL.IISC.ERNET.IN>
Subject: Chaos and Fractals in Hydrology
To: Multiple recipients of list CIVIL-L <CIVIL-L@unb.ca>
Hi Civilians,
My team here had a sort of disappointing experience with the neural
network stuff (applied to hydrologic time series) - No, the results were
quite encouraging and perhaps the work may bring us a couple of
not-so-bad publications too, but there was no excitement, no creativity
(unless you call the late-night hours on the computer, mechanically going
through the training of the network with different parameter sets as
creativity), not even the frustrations of initial failures. And that is
what made the experience a trifle dull.
Is the "chaos and fractals" stuff similar to the neural nets.. or is
there any modelling excersise involved in it that calls for a good deal
of background and experience in hydrology, a bit of creativity and
perhaps some excitement.
Should love to interact on this with someone working in
this area.