Research introduction:
I have a broad variety of interests in research. My overall research focuses
on the area of machine learning and data mining, but fundamentally I am
a bit of a computer statistician. Projects involving software or software
agents that learns to perform better interest me.
Suggested project areas
- Robotic gaming - I am very interested in Robotic games, and
especially the RoboCup simulated soccer game. In previous research my
students have developed a basic interface for working with RoboCup, I would
be interested in developing learning agents working with that interface
to try to learn how to play the game.
- Sports statistics - the huge growth in sports statistics and our
better understanding of how those statistics help us better understand
performance is highly interesting to me. Though there has been a lot of
work on baseball statistics, less has been done with statistics in other
sports (for example, basketball is just seeing the introduction of ideas
such as tempo-free offense). I would be interested in working on methods to
crawl the web for data (extracting box scores for example) and on predictive
methods.
- Bioinformatics - I have a project for implementing a known
algorithm for bi-clustering data (basically clustering data together in
groups, but where you consider two different types of labels for the groups).
This project could lead to a publication if it works well and we would likely
make the software available on the web.
Contact me - email rmaclin@d.umn.edu to get an appointment