Short Vita:

Evangelos Theodorou earned his bachelor in Electrical and Computer Engineering form the Technical University of Crete (Greece) at 2001. During the last two years of his undergrad studies, he worked as machine vision engineer for European research projects on industrial vision. In 2003, he got his master's in industrial engineering from the Technical University of Crete. In his graduate studies back in Greece he was working as research assistant in the Dynamic Systems and Simulation Laboratory(DSSL). In his work in DSSL he designed, tested and evaluated stochastic optimal controllers applied to traffic urban networks. In 2004 he was accepted by the Computer Science and Engineering Department in the University of Minnesota USA. During his graduate studies in Minnesota he was research assistant in the Computational Perception and Action Lab (CPAB) and in 2007 he graduated with his 2nd masters. In may 2010 he graduated with masters in controls from Electrical Engineering dept at USC and in May 2011 with his PhD in Computer Science Dept at USC. His doctoral research was performed in Computational Learning and Motor Control Lab in Computer Sciene Dept and the Brain and Body Dynamics Lab (BBD) in the Department of Biomedical Engineering under the supervision of Dr. Stefan Schaal (main advisor) and Dr. Francisco J-Valero Cuevas (Co-advisor). Currently he is holding a postdoctoral research associate position in University of Washington Seattle with the department of Computer Science and Engineering and he is working under the supervision of Dr. Emo Todorov in the movement control laboratory (MCL).

Here is my CV: PDF

Journal Publications :

[1] Rombocas E, Mathora M, Theodorou E and Todorov E, Matsuoka Y. Reinforcement Learning and Synergistic Control of the ACT hand. UNDER REVIEW

[2] Zhe Xu, Kumar V, Theodorou E, Matsuoka Y and Todorov E. A Compliant Biomimetic Artificial Finger for Anthropomorphic Robotic Hands via 3D Rapid Proto- typing. UNDER REVIEW

[3] Stulp F, Theodorou E and Schaal S Hierarchical Reinforcement Learning for Robust Manipulation. Submitted to IEEE TRANSACTIONS ON ROBOTICS AND AUTOMATION CONDITIONAL ACCEPTANCE

[4] Theodorou E, Valero-Cuevas FJ and Schaal S Understanding Tendon Coordination for Index Finger Movements: Analysis and Model Verification with Path integral Reinforcement Learning. UNDER REVIEW

[5] Mistry M, Theodorou E, Schaal S, and Kawato M. Adaptation to a suboptimal desired trajectory. UNDER REVIEW

[6] Theodorou, E. A. (2011). Iterative path integral stochastic optimal control: Theory and applications to motor control. PhD Thesis, University of Southern California. PDF

[7] Buchli J, Stulp F, Theodorou E and Schaal S. (2011). Learning variable impedance control, International Journal of Robotics Research. PDF

[8] Theodorou E, Buchli J, and Schaal S (2010). A Generalized Path Integral Control Approach to Reinforcement Learning, Journal of Machine Learning Research,11, pp.3137-3181 PDF Erratum: PDF

[9] Valero-Cuevas FJ, Hoffmann H, Kurse M, Kutch JJ, Theodorou E(2009). Computational models for neuromuscular function, IEEE Reviews in Biomedical Engineering --(All authors have equally contributed), 2, pp.110 -135. PDF

Conference Publications:

[1] Theodorou E, Todorov E. Girsanov Based Direct Policy Gradient Methods. PDF, Manuscript Under Review

[2] Theodorou E, Todorov E. Relative Entropy Free Energy Dualities: Connection to Path Integral and KL control. PDF, Manuscript Under Review

[3] Theodorou E, Todorov E. Stochastic Optimal Control of Nonlinear Markov Jump Diffusion Processes. Accepted in American Control Conference (ACC) 2012. PDF

[4] Malhotra M, Rombokas E, Theodorou E, Todorov E and Matsuoka Y, Reduced dimensionality control for the ACT hand. Accepted to the International Conference of Robotics and Automation (ICRA) 2012. PDF

[5] Rombokas E, Theodorou E, Malhotra M, Todorov E and Matsuoka Y. Tendon-driven control of biomechanical and robotic systems: A path-integral reinforcement learning approach.
Accepted to the International Conference of Robotics and Automation (ICRA) 2012. PDF

[5] Meyer F, Theodorou E, Schaal S. Movement Segmentation and Recognition for Imitation Learning. Accepted to AISTATS 2012.

[6] Stulp F, Theodorou E, Schaal S. Learning Motion Primitive Goals for Robust Manipulations. In the International Conference on Intelligent Robotic Systems, (IROS) 2011.PDF

[7] Theodorou E, Stulp F, Schaal S. Path Integral Reinforcement Learning. In the Proceedings of the 15th Yale Workshop on Adaptive and Learning Systems, 2011. PDF

[8] Meyer F, Theodorou E, Stulp F, Buchli J, Schaal S. Movement segmentation using a library of primitives. In the International Conference on Intelligent Robotic Systems, San Francisco, (IROS) 2011.PDF

[9] Stulp F, Buchli J, Ellmer A, Mistry M, Theodorou E, Schaal S. Reinforcement Learning of Impedance Control in Stochastic Force Fields. In IEEE International Conference on Development and Learning,
and Epigenetic Robotics, Frakfurt, Germany, (ICDL) 2011.

[10] Theodorou E, Stulp F, Buchli J and Schaal S. Iterative Path Integral Stochastic Optimal Control for Learning Robotic Tasks. In the 18th World Congress of The International Federation of Automatic Control, Milan Italy, (IFAC)2011. PDF

[11] Daniel A. Braun, Pedro A. Ortega, Theodorou E, and Schaal S. Path Integral Control and Bounded Rationality. In IEEE conference of Approximate Dynamic Programming and Reinforcement Learning, Paris, (ADPRL) 2011. PDF

[12] Pastor P, Kalakrishnan M, Chitta S, Theodorou E, Schaal S. Skill Learning and Performance Prediction for Manipulation. In IEEE International Conference of Robotics and Automation, Shanghai, China, (ICRA) 2011.PDF

[13] Kalakrishnan M, Chitta S, Theodorou E, Pastor P, Schaal S. STOMP: Stochastic Trajectory Optimization for Motion Planning. In IEEE International Con- ference of Robotics and Automation, Shanghai, China, (ICRA) 2011. PDF

[14] Stulp F, Theodorou E, Buchli J, Schaal S. Learning to Grasp under Uncertainty. In IEEE International Conference of Robotics and Automation, Shanghai, China (ICRA) 2011. PDF

[15] Theodorou E, Todorov E, and Valero-Cuevas FJ. Neuromuscular Stochas- tic Optimal Control of a Tendon Driven Index Finger. American Control Conference, San Francisco, (ACC) 2011.PDF

[16] Stulp F, Buchli J, Theodorou E and Schaal S. Reinforcement Learning of Full- body Humanoid Motor Skills. In International Conference on Humanoid Robotics 2010. Finalist for the best paper award.

[17] Theodorou E and Valero-Cuevas FJ. Optimality in Neuromuscular Systems. In 32nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society, (EMBS) 2010.

[18] Buchli J, Theodorou E, Stulp F, and Schaal S. Variable Impedance Control - A Reinforcement Learning Approach . In Robotic Systems Sciences (RSS) 2010. PDF

[19] Theodorou E, Buchli J, Freek S, Schaal S. An Iterative Path Integral Reinforcement Learning Approach. Sbowbird Learning Workshop, 2010. PDF

[20] Theodorou E, Todorov E, and Valero-Cuevas FJ. A First optimal control solution for complex nonlinear, tendon driven neuromuscular Finger model. In American Society of Mechanical Engineering, May, (ASME) 2010. PDF

[21] Theodorou E, Buchli J, and Schaal S. Learning Policy Improvement with Path Integrals. In International Conference on Artificial intelligence and Statistics (AISTATS) 2010.

[22] Theodorou E, Tassa Y and Todorov E. Stochastic Differential Dynamic Programming. In the Proceedings of American Control Conference (ACC), 2010. PDF

[23] Theodorou E, Buchli J, and Schaal S. Path- Integral Stochastic Optimal Control for Rigid Body Dynamcis. In IEEE symposium on Adaptive Dynamic Programming and Reinforcement Learning, (ADPRL) 2009. PDF

[24] Ting J, Theodorou E, and Schaal S. Learning an Outlier - Robust Kalman Filter. In European Conference on Machine Learning (ECML), 2007 PDF

[25] Ting J, Theodorou E , and Schaal S. Kalman Filter for Robust Outlier Detection. In IEEE International Conference on Intelligent Robotic Systems (IROS), 2007. PDF

Peer Review Abstracts:

[1] Hoffmann H, Theodorou E, and Schaal S. Optimization Strategies in Human Reinforcement Learning. In: Advances in Computational Motor Control VII, Sympo- sium at the Society of Neuroscience Meeting, Washington DC, 2008.

[2] Mistry M, Theodorou E, Liaw G, Yoshioka T, Schaal S, and Kawato M. Adap- tation to a suboptimal desired trajectory. In: Advances in Computational Motor Con- trol VII, Symposium at the Society of Neuroscience Meeting, Washington DC, 2008.

Abstracts:

[1] Theodorou E, and Schaal S. Learning Optimal Control Solutions: A Path Inte- gral Approach. In: Neural Control of Movement Conference, May, 2010.

[2] Hoffmann H, Theodorou E, and Schaal S. Human optimization strategies under reward feedback. In: Neural Control of Movement Conference, May, 2009. Poster abstract.

[3] Mistry M, Theodorou E, Hoffmann H, and Schaal S. The dual role of uncertainty in force field learning experiments. In: Neural Control of Movement Conference, May, 2008. Poster abstract.

[4] Hoffmann H, Theodorou E, and Schaal S. Behavioral Experiments on reinforce- ment learning in human motor control. In : Neural Control of Movement Conference, May, 2008. Poster abstract.

[5]. Mistry M, Theodorou E, Schaal S, and Kawato M. Uncertain 3D force field in reaching movements: Do humans favor robust or average performance. In: Society of Neuroscience meeting, 2007. Poster abstract.

[6]. Theodorou E, Peters J, and Schaal S. Policy Gradient methods for optimal control of arm Movements. In: Society of Neuroscience meeting, 2007. Poster abstract.

Technical Reports:

[1] Ting J, Theodorou E, Schaal S (2007). Learning an Outlier-Robust Kalman Filter, CLMC Technical Report: TR-CLMC-2007-1.

[2] Theodorou E and Schrater P. Statistical Learning of LQG controllers. Technical Report: UMN - TR - 2006 - 1.

[3] Theodorou E, Linear and Nonlinear Estimation models applied to Hemodynamic Model. Technical Report: UMN: TR - 2005 - 1.

[4] Theodorou E and Hidaka Y. Parametric and Nonparametric approaches to Tracking of moving objects. Technical Report: UMN- TR - 2005 - 2.

Evangelos A. Theodorou