Support System for Environmental Decision Making
This project started several years ago. Initially, it involved the use of
machine learning programs to model an existing algorithm for selecting
between mathematical models for simulating groundwater contaminant
transport phenomena. The algorithm has been developed at a part of a
comprehensive advisory system since 1988 by
Prof. Medina and his
colleagues at
Duke University. Since then, the algorithm has been deployed as part of a
comprehensive advisory system in the Ground Water Section, State of North
Carolina, the U.S. Air Force through the Armstrong Laboratory, Tyndall,
ABF, FL, and the U.S. Army Corp of Engineers Waterways Experiment Station,
MI.
Through the use of machine learning programs such as IND, CN2, and FOIL, we
were able to model the algorithm and debug it. Since then, we have used the
same procedure to model improved versions of the algorithm. The results of
this study are summarized in:
- Reich, Y., Medina, M. Jr., Shieh,
T.-Y., and Jacobs, T. (1996), Modeling and Debuging
Engineering Decision Procedures with Machine Learning, Journal of
Computing in Civil Engineering,10(2):157-166
(Postscript
file, 290K)
To read more about the advisory system and other topics related to
contaminant transport, infrastructure rehabilitation, site remediation, and
risk assessment related to groundwater, go to the Laboratory for
Contaminant Hydrology Systems Simulation and Decision Making ,
Duke University (page under construction).
Presently, we are expanding the scope of the study to design a
comprehensive environmental decision making that will take into account
many more issues. More about this research will be available here in the
future.
If you have any comments or questions, please send
me a message.
Last modified by Yoram Reich on September 15, 1995