Michael T. Rosenstein Senior Research Scientist with the Laboratory for Perceptual Robotics and the Autonomous Learning Laboratory
Department of Computer Science University of Massachusetts 140 Governors Drive Amherst, MA 01003-9264
NOTE: This web site is for archival purposes. I am currently a research scientist at iRobot.

Education
Univ. of Massachusetts Amherst		Ph.D. in Computer Science		May, 2003
Boston University		M.S. in Biomedical Engineering		January, 1993
Boston University		B.S. in Applied Physiology		May, 1989

Research Interests
My research is multidisciplinary, involving aspects of machine learning, robotics, and human behavior. In broad terms, my goal is to build intelligent computer systems that regularly interact with people. Specific areas of interest include the following: Transfer learning, reinforcement learning, and learning from demonstration Human-computer interaction and haptic user interfaces Telerobotics and supervisory control of prosthetic/orthotic devices Modeling of physiological activity, including postural control and human gait Human motor development, coordination, and optimization as a principle of motor learning
Selected recent projects: Human-Robot Interaction and Predictive Displays NASA's Robonaut Supervised Actor-Critic Reinforcement Learning Selected past projects: Robot Weightlifting and Motor Synergies Dynamic Manipulability Nonlinear Time Series Analysis Sensorimotor Categorization Biped Locomotion

Selected Publications [complete list]
Supervised actor-critic reinforcement learning
	M.T. Rosenstein and A.G. Barto. In J. Si, A. Barto, W. Powell, and D. Wunsch, eds., Learning and Approximate Dynamic Programming: Scaling Up to the Real World. John Wiley & Sons, Inc., New York, 359-380, 2004. [pdf] [ps.gz]
Learning to Exploit Dynamics for Robot Motor Coordination
	Ph.D. thesis, University of Massachusetts Amherst, 2003. Committee: Andrew G. Barto (chair), Neil E. Berthier, Andrew H. Fagg, Roderic A. Grupen and Richard E. A. Van Emmerik. [pdf]
Velocity-dependent dynamic manipulability
	M.T. Rosenstein and R.A. Grupen. In Proceedings of the IEEE International Conference on Robotics and Automation, vol. 3, 2424-2429, 2002. [pdf] [ps.gz]
Robot weightlifting by direct policy search
	M.T. Rosenstein and A.G. Barto. In Proceedings of the Seventeenth International Joint Conference on Artificial Intelligence, vol. 2, 839-844, 2001. [pdf] [ps.gz]
Continuous categories for a mobile robot
	M.T. Rosenstein and P.R. Cohen. In Proceedings of the Sixteenth National Conference on Artificial Intelligence, 634-640, 1999. [pdf] [ps.gz]
Reconstruction expansion as a geometry-based framework for choosing proper delay times
	M.T. Rosenstein, J.J. Collins, and C.J. De Luca. Physica D 73:82-98, 1994. [pdf] [ps.gz]

Selected Awards and Honors
2001, 2002	NASA Graduate Student Researchers Program Fellowship (one-year tenure, renewable)
1997	National Defense Science and Engineering Graduate Fellowship (three-year tenure)

Teaching
EXCSCI 897V	Bernstein's Contributions to Motor Control	Spring, 2001
EXCSCI 797N	Nonlinear Dynamics of Human Movement	Spring, 2000
EXCSCI 697N	Nonlinear Biodynamics	Spring, 1999

Miscellany
Latest robotics news
The AAAI-04 Workshop on Supervisory Control of Learning and Adaptive Systems
Schedule of conferences on artificial intelligence, robotics and motor control
L1D2
	A tool for estimating the largest Lyapunov exponent and correlation dimension from a time series. Executables and source code are available from the PhysioNet web site.

updated 07-Aug-2006