Inside this Issue:

Download the Newsletter

Editors:

Joseph Cummings
Charles Dizard

Informatics Today is a publication of the University of Pittsburgh Department of Biomedical Informatics

The University of Pittsburgh is an affirmative action, equal opportunity institution. Published in cooperation with the Department of University Marketing Communications.

UMC65727-1008

Featured Alumni

Q&A with Denver Dash,
PhD, 2003, Post Doctoral Research Fellow
Research Scientist, Intel Research Pittsburgh

Question: What have you been doing since completing your postdoctoral fellowship at Pitt?
Answer: I did a brief fellowship with Greg Cooper at the Center for Biomedical Informatics before joining Intel Research in Santa Clara, Calif. as a research scientist in late 2003. I moved to the Intel Research Pittsburgh lab on the campus of Carnegie Mellon University in 2006. At Intel, I've been applying graphical models to detecting intrusion detections in networks, and lately I've been working on improving computer perception by incorporating many diverse sensors and multiple perspectives into the problem and on parallelizing machine learning algorithms.

Question: What are your most memorable experiences from your days at Pitt?
Answer: I enjoyed summers at Lake Arthur, spelunking, Latin dancing at Rosebud on Thursdays, playing ultimate Frisbee, teaching a semester in Poland, and meeting and courting my amazing wife.

Question: How did your training and education benefit your career?
Answer: My advisor, Marek Druzdzel, did a great job preparing me for my career. He gave me a solid foundation in Bayesian networks with his decision support systems class. He exposed me to excellent programmers who taught me a lot about first-class development, and he encouraged me to teach a course in Poland which forced me to polish my knowledge of machine learning well enough to teach it to others. Greg Cooper's reading courses in bioinformatics gave me a great broad-based background in that field that has been invaluable as well. Kirk Pruhs' algorithms class from the Pitt computer science department has been one of those classes that I am grateful for every day.

Question: What are your research interests?
Answer: I have been pretty active in disease outbreak detection research, much of which can be applied to intrusion detection as well. Recently, I have been working on cluster onset detection, a method for real-time unsupervised clustering to detect the onset of possibly diseased clusters in a population. I'm also very interested in taking a lot of these machine learning and detection algorithms and understanding how to implement them efficiently in a cluster-computing environment. A lot of people don't realize that Moore's law on a single processor is basically over. If we want our algorithms to continue to exploit the benefits of microprocessor advances, the only way forward is to look at parallel and distributed computing methods.

Question: What advice would you give to current fellows in the training program?
Answer: If you expect to be doing any programming, I would strongly recommend taking Pruhs' algorithms class in the computer science department. It has a lot of broad applicability, and I have personally found it invaluable in my career. I would also recommend looking around at other departments and taking advantage of the strong interdisciplinary atmosphere that being at Pitt provides. I would include in that some of the great offerings available via cross-registration at Carnegie Mellon University.

Question: Anything about your personal life you would like to share?
Answer: A few months ago, my wife gave birth to our first child, Dagny Dash. Even in such a short time I have been amazed and inspired by how human beings are able to perceive the world around them and become self-aware, not to mention store nearly an entire lifetime's worth of experience and memories in such a compact form-factor.