http:^^www.cs.unm.edu^~lisa^

This data set contains WWW-pages collected from computer science departments of various universities

Date: Tue, 26 Nov 1996 17:50:04 GMT
Server: NCSA/1.4.2
Content-type: text/html
Last-modified: Thu, 23 May 1996 05:58:27 GMT
Content-length: 2349

<TITLE>  Desjarlais Home Page </TITLE>
<H2> 
Graduate Student: Lisa Maynes Desjarlais </H2>
<HR>

I obtained a B.A. in Physics from 
<!WA0><A HREF = "http://www.cornell.edu"> Cornell University</a>
in 1981, and an M.S. in Computer Graphics from 
<!WA1><A HREF = "http://www.graphics.cornell.edu"> Cornell University's Program of
Computer Graphics</a> in 1986.
I am currently working on a Ph.D. in Computer Science at the
<!WA2><A HREF = http://www.cs.unm.edu/> <B> University of New Mexico</B></A> 
in the area of genetic algorithm-based machine learning.
My advisor is
<!WA3><A HREF = http://www.cs.unm.edu/~forrest/>
Stephanie Forrest</A>.


<HR>

<H3> 
<!WA4><img align=bottom src="http://www.cs.unm.edu/research.gif">
Recent Research Projects
</H3>

I am looking at ways to combine two forms of genetic algorithm-based 
machine learning--learning classifier systems (LCSs) and 
<!WA5><A HREF = http://www.salford.ac.uk/docs/depts/eee/gp2faq.html>
<B> genetic programming </B></a>
(GP)--in hope of remediating the limitations of one 
approach with the abilities of the other.

I conduct my research at the 
<!WA6><A HREF = http://www.sandia.gov/cc_at/lcc_at.html/> 
<B> Intelligent Systems and Robotics Center</B></a>
at <!WA7><A HREF = http://www.sandia.gov/> 
<B> Sandia National Laboratories</B></a>.
For them, I am also developing a simulator for large numbers of 
autonomous agents moving within and interacting with a physical environment. 
The agents are tiny robots, or "robugs", which we will eventually really 
build. 
In the current version of the simulator, geometric and physical aspects of 
the simulation, such as visualization and sensor simulation, are implemented 
using SGI's Performer package. I am currently trying to implement my first
sensor, a whisker, using Performer's line segment intersection capability.
When we've got the basic physical simulation going, our emphasis will
shift to effecting, testing, and automatically generating robug behavior,
first of individual bugs and eventually the collective behavior of
swarms of many bugs. The robug simulator will become the
testbed for the learning algorithms I'm developing for my thesis.

<HR>
<!WA8><img align=top src="http://www.cs.unm.edu/subpages/UNMSEALc.gif">

<ADDRESS> 
<!WA9><A HREF = http://www.cs.unm.edu/cgi-bin/finger?lisa@chaco>
[e-mail]:          [phone]
<BR>
[e-mail]:  [phone]
</a></ADDRESS>