Welcome to my home away from home!
Here's a link to My Old Homepage.
|
Lucas Shaw
Graduate Student |
|
![]() |
812 E. University Ave. Laramie, Wyoming 82072
|
| Education: |
|
||||||||||||
| Advisor: | Associate Professor Dr. William D. Spears | ||||||||||||
| Research Interests: |
I am still working on a parallelized evolutionary algorithm framework (PEAF) using the LAM-MPI implementation of the Message Passing Interface (MPI). This project is designed to supplement my simulation research as well as to provide a tool for scientists to develop their own genetic algorithms, evolution strategies, or evolutionary programs to aid them in understanding their complex problems. It has turned out to be an interesting tool, and we are testing the effectiveness of using heterogeneous ensembles of EAs to work on a problem. This has turned out to be a really good general approach to searching a search space. We are still testing various aspects of the EAs we have developed with the framework. I am also still working with my flu virus spread simulation. Last semester, I verified that my C++ and starlogo versions are equivalent. Dr. Spears and I extended the mathematics of the original simulation to take into account the modeling of the vaccinations. We also derived some properties of the simulation that will be useful in guiding my exploration of the simulation, its uses, flaws, etc. We tested many of the properties we devised, though that is ongoing. I still want to add quarantine procedures to it, but right now, I am trying to make it more efficient, since it takes roughly twice as long to run on the Beowulf cluster as it does on Dr. Spear's super-duper dual-processor computer (communication costs are killing me). I also need to figure out how to weave my simulation into a fitness function that will run in my PEAF (more on that in Current Projects). |
||||||||||||
| Current Projects: |
I am continuing my work on the PEAF (Parallelized Evolutionary Algorithms Framework). I have implemented a few PEAs (Parallel Evolutionary Algorithms). I am also trying to get my thesis going. I failed to implement a PEA to solve a resource allocation problem with my flu virus simulation for the Distributed Agents class and my thesis, so I will try to adapt PEAF and my simulation so that they work together better (and develop a technique for using parallelized fitness functions in my framework). Along those lines, I plan on refactoring my FEAF so that it will be configurable from a configuration file (for user friendliness). All these plans depend on what direction my research needs to go for me to finish up with everything this semester (The ultimate goal is graduation in May). I am enrolled in Research Writing, which promises to be a challenging and interesting course. |
||||||||||||
| Current Courses: |
| ||||||||||||
| Old Courses: |
| ||||||||||||
| Old Projects: |
I am continuing my work on the PEAF (Parallelized Evolutionary Algorithms Framework). PEAF uses the LAM implementation of the Message Passing Interface (MPI) as the vehicle of parallelization. This project is designed to supplement my simulation research as well as to provide a tool for scientists to develop their own genetic algorithms, evolution strategies, or evolutionary programs to aid them in understanding the complex problems. I have implemented a few PEAs (Parallel Evolutionary Algorithms), one to solve the TSP (Travelling Salesperson Problem), another to minimize a generalized form of the Rosenbrock function, and another CNF XSAT (Conjunctive Normal Form Satisfiability) -- to solve 3SAT problems. I failed to implement a scalable PEA to solve a resource allocation problem with my flu virus simulation for the Distributed Agents class (and my thesis). I ran into a problem with too many file pointers being used and the process limit of the LAM-MPI implementation (since my simulation uses 13 processes per instance by itself). I pushed solving those problems back, and will address them as I have time. I derived some equilibrium behavior of the vaccine addition to the virus simulation, then I verified that the simulation agreed with the results. In COSC-5840 Software Engineering, my group created a Firefox Extension that displays RSS feeds. We called it Hunter and it is available for download here (http://www.cs.uwyo.edu/~lucshaw/SE).
I am still trying to understand my simulation and how it may be used practically. A vaccination capability has been added to the simulation. I need to add more powerful vaccination capabilities as well as quarantine procedures as well. I worked out the parallel EA framework, which I called PEAF. It is up and running with TSP and Rosenbrock functions as the test bed. I was also working on the network engine for the Hidden Agenda game Wes, Glen, and I were developing. The network engine project got canned, so Wes had to do the game by himself. The game turned out to be quite good and was the most Technically Advanced game in the competition. Congratulations to Wes for that! In 5850, we worked on a Research Portal and Idea Sharing Web Service for COSC using Java, jsp's, and Tomcat. In 5350 System Simulation, I created a simple simulation of a small two person post office. In COSC-5700 Computer Architecture, I worked with Huiyuan Ma on optimizing a determinate calculation for Professor Chtcherba. |
||||||||||||
| Other links: |
Handy Shell Scripts CHS Class of 1996 Fire at the Bull Elk Park Fire, 1999 My fellow Fire Fighter Type II crewmates I loved to stand on hy head when I was young. Drawing 1 Lawn Art, 1998 My cat is possessed! My Chilli Recipe *coming soon |