DM: datamine-l new member intro

[William Hsu]

Hello,

        Here's my belated intro to the DM group:

        My name is William Hsu.  I am a Ph.D. student at the 
University
of Illinois at Urbana-Champaign, and work in the Beckman Institute AI
Group.  I'm currently completing my dissertation on spatiotemporal 
sequence
learning (specifically, causal discovering and data mining from 
spatially
reference time series).  Some applications of this research are 
situated
crisis recognition (critiquing, control systems for automation in 
crisis
management) and monitoring.  My primary test bed is a geospatial 
decision
support problem in precision agriculture and applied climatology.  
It's
of interest to me primarily because of its heterogeneity (e.g., it
contains many different embedded autoregressive, moving average, ARMA,
duration-based time series modeling subproblems, among others) and
decomposability.

        I'm also looking at smaller, more homogeneous test beds in the
area of biomedical signals, protein secondary fold prediction, and
smaller-scale time series monitoring.

        The methods I am particularly interested in are: model 
selection,
Bayesian ensemble learning (variational techniques, Markov chain Monte
Carlo, etc.), (hierarchical) decomposition of inductive learning 
tasks,
hybridization of ANN and Bayesian network learning, modular 
connectionist
(graphical, probabilistic) models, and integration of symbolic and
seminumerical learning.

-Bill

P.S. : I have been lurking for a couple of months but am very 
interested
       to know if anyone else is working on technique selection from
       a ``repertoire'' of probabilistic models.  In my case, the 
toolbox
       is populated by temporal ANNs and Bayesian networks.

P.P.S. : I've been trying to gather interest/momentum for a USENET 
newsgroup,
         comp.ai.uncertainty, to discuss these and other uncertain 
reasoning
         (and KDD) topics.  Please let me know if you have any such
         interests!

------------------------------------------------------------------------------
 >> Visit the UIUC Neural Networks Home Page at 
 >http://anncbt.ai.uiuc.edu <<
------------------------------------------------------------------------------
     >>>>>  Hosted by the Knowledge Based Systems (UIUC-KBS) Lab <<<<<
*|/////||==============- http://www-kbs.ai.uiuc.edu 
-==============||\\\\\|*
------------------------------------------------------------------------------