Carla Gomes

Research Associate
gomes@cs.cornell.edu
http://www.cs.cornell.edu/gomes
Ph.D. University of Edinburgh, 1993

My research interests are centered around the integration of methods from artificial intelligence and operations research for applications in planning and scheduling, and, more generally, multidisciplinary approaches for solving combinatorial problems. Recently, I have focused on randomized search techniques. In this work, I study so-

called heavy-tailed distributions that characterize complete randomized search methods. A promising way of exploiting heavy-tailed behavior is by combining a suite of search methods into a portfolio, running on a distributed compute cluster. It can be shown that such portfolios dramatically reduce the expected overall computational cost, thereby allowing us to solve large, previously unsolved planning and scheduling problems. Using a recent equipment grant, we are constructing a 32-node parallel compute cluster with a high-speed communication network to further evaluate and study our algorithm portfolio approach.

Honors and Awards

Best Consultant, Information Directorate Air Force Research Laboratory (AFRL).

Professional Activities

Guest Editor: Journal of Knowledge Engineering Review, Cambridge Press.

Editorial Board Member: Journal of Knowledge Engineering Review.

Co-Chair: AAAI Workshop on Leveraging Probability and Uncertainty in Computation, AAAI 2000.

Member: Advisory Committee International Scientists, Ministry of Science and Technology, Portuguese Government, Presidency of European Union, 2000; DARPA Information Science and Technology Study Group on Probabilistic Methods in Computational Systems and Infrastructure; DARPA Information Science and Technology Study Group on Self-Configuring Wireless Sensor Networks; NASA External Review Panel.

Reviewer: Journal of Automated Reasoning; Journal of Artificial Intelligence Research; Constraints: An International Journal; Discrete Applied Mathematics.

Lectures

Structure, Duality, and Randomization—Common Themes in AI and OR. Invited plenary talk. 17th Natl. Conf. on Artificial Intelligence (AAAI-00).

Heavy-Tailed Phenomena in Computation. DARPA Information Science and Technology Study Group on Self-Configuring Wireless Sensor Networks, Berkeley, CA, March 2000.

Randomization in Computation. Distinguished Lecture Series, University of Lisbon, Lisbon, January 2000. Structure, Duality, and Randomization. New World Vistas AFOSR Principal Investigator Meeting, Florida, 2000.

Leveraging Randomization in Computation. Scientific Advisory Board, Information Directorate, AFRL, December 1999.

Publications

“Artificial intelligence and operations research: challenges and opportunities in planning and scheduling and operations research.” Journal of Knowledge Engineering Review 15 (1) (2000).

“Algorithm Portfolios.” Artificial Intelligence Journal 2000 (with B. Selman).

“Heavy-tailed phenomena in satisfiability and constraint satisfaction problems.” Journal of Automated Reasoning 24 (1/2) (2000), 67–100 (with B. Selman, N. Crato, and H. Kautz). Results featured in Science News, May 2000.

“Hybrid strategies for heterogeneous search spaces.” International Journal on Artificial Intelligence Tools (with B. Selman).

“Structure, duality, and randomization - common themes in AI and OR.” Proc. of the 17th Natl. Conf. on Artificial Intelligence (AAAI-00), 2000.

“Generating satisfiable instances.” Proc. of the 17th National Conference on Artificial Intelligence (AAAI-00), 2000 (with D. Achlipotas, H. Kautz, and B. Selman).

“On the fine structure of large search spaces.” Proc. of the International Conference on Tools Artificial Intelligence (ICTAI-99), Chicago (1999) (with B. Selman).

“Search strategies for hybrid search spaces.” Proc. of the International Conference on Tools Artificial Intelligence (ICTAI-99), Chicago (1999) (with B. Selman).