Computer Science Assistant

Location:

Posted:

November 04, 2012

Philip S. Thomas

*******@**.*****.***

www.psthomas.com

Education University of Massachusetts Amherst, Amherst, Massachusetts

Ph.D. Candidate, Computer Science, in progress, GPA: 3.94

Adviser: Professor Andrew G. Barto

Case Western Reserve University, Cleveland, Ohio

Master of Science, Computer Science, August 2009

Bachelor of Science, Computer Science, May 2008, Cum Laude

Minor in Artificial Intelligence and Psychology

Publications P.S. Thomas and A.G. Barto. Conjugate Markov decision processes. To appear, Proceedings of the

Twenty-Eighth International Conference on Machine Learning, 2011.

G. D. Konidaris, S. Osentoski, and P.S. Thomas. Value function approximation in reinforcement

learning using the Fourier basis. To appear, Proceedings of the Twenty-Fifth Conference on

Artificial Intelligence, August 2011.

P.S. Thomas, A Reinforcement Learning Controller for Functional Electrical Stimulation

of a Human Arm, M.S. Thesis, Electrical Engineering and Computer Science Department.,

Case Western Reserve University, Cleveland, OH, August 2009.

Advisor: Professor Michael S. Branicky

P.S. Thomas, M.S. Branicky, A. van den Bogert, K.M. Jagodnik. Application of the actor-critic

architecture to functional electrical stimulation control of a human arm. Proceedings of the Twenty-

First Innovative Applications of Artificial Intelligence, 2009.

P.S. Thomas, M.S. Branicky, A. van den Bogert, K.M. Jagodnik. Creating a reinforcement learning

controller for functional electrical stimulation of a human arm.

Proc. Fourteenth Yale Workshop on Adaptive and Learning Systems, 2008.

Research Intelligent Control of Upper Extremity Neural Prostheses, (2006-2009)

Experience Research Assistant, under NIH Grant R21HD049662, developing a reinforcement learning

controller for functional electrical stimulation (FES) control of a paralyzed human arm.

DARPA Urban Challenge 2007, Team Case team member (2006-2007)

Developed and implemented constrained multi-dimensional planner. Result was a heavily filtered

cross between a Feasibility Graph and a weighted, bidirectional, nonholonomic, geometrically

constrained Rapidly Exploring Random Tree, with sub-trees grown from choke points in the

configuration space. Final version was run on Dexter at the 2007 Urban Challenge NQE.

Teaching Teaching Assistant, (Fall 2009 - Spring 2011, 4 semesters)

Experience Teaching Assistant, for University of Massachusetts Amherst, course CMPSCI 311: Introduction to

Algorithms.

Graduate Assistant, (Summer 2010)

I assisted with NSF Research Experience for Undergraduates (REU) program run by the

Department of Computer Science at the University of Massachusetts Amherst. I ran several

workshops to bridge the knowledge gap between undergraduate coursework and graduate research.

Honors Honorable Mention, NSF Graduate Research Fellowship Program (2009)