Ph.D. Student Wins Award at ICML 2006

July 4, 2006

ATLANTA (July 5, 2006)--College of Computing Ph.D. student Michael Holmes was awarded the "Distinguished Student Paper" Award at the 23rd International Conference on Machine Learning (ICML). ICML is one of the premiere machine learning conferences, and is the largest international conference on machine learning research and applications. Holmes received the award for the paper he wrote with Assistant Professor and Advisor Charles Isbell titled: Looping Suffix Tree-Based Inference of Partially Observable Hidden State.

The paper provides an algorithm for inferring hidden state from sensorimotor experience when the environment takes the form of a POMDP with deterministic transition and observation functions. Such environments can appear to be arbitrarily complex and non-deterministic on the surface, but are actually deterministic with respect to the unobserved underly- ing state. The paper shows that there always exists a finite history-based representation that fully captures the unobserved world state, allowing for perfect prediction of action effects.

Holmes is in his fourth year as a Ph.D. student within the College's Interactive and Intelligent Computing (IIC) division. He received his BS degrees in Physics and Computer Science, with a minor in Math. The 2006 ICML was held June 25-29 in Pittsburgh, PA.

To view Holmes' and Isbell's award-winning paper, click here.