Scalable Statistical Learning for Robotics

We are interested in supervised learning methods that accomplish nonlinear coordinate transformations and achieve robust internal models for our autonomous high-dimensional anthropomorphic systems.

Our focus is on the development of new learning algorithms for complex movement systems, where learning may proceed in an incremental fashion (i.e. sequential availability of data points). Using Bayesian approaches and graphical models, we aim to create algorithms that are fast, robust and based on a solid statistical foundation, yet scalable to extremely high dimensions.

Recent work has included the Bayesian Backfitting Relevance Vector Machine and a related variant (applied to EMG activity prediction). Both produce computationally efficient solutions and offer properties such as feature detection and automatic relevance determination. An augmented version of the algorithm's graphical model gives a Bayesian Factor Analysis regression model that performs noise cleanup. It offers significant improvements in generalization performance, as has been demonstrated in parameter identification tasks for our robotic platforms.

Contact persons: Jo-Anne Ting, Stefan Schaal

Publications

Ting, J.; Theodorou, E.; Schaal, S. (2007). Learning an Outlier-Robust Kalman Filter, European Conference on Machine Learning (ECML 2007), pp.748-756, Springer.
[Keywords: automatic outlier detection, kalman filter, system dynamics, weighted least squares, bayesian statistical learning]
[Detail] [BibTeX] [PDF]

Ting, J.; Theodorou, E.; Schaal, S. (2007). Learning an Outlier-Robust Kalman Filter, CLMC Technical Report: TR-CLMC-2007-1.
[Keywords: automatic outlier detection, kalman filter, system dynamics, weighted least squares, bayesian statistical learning]
[Detail] [BibTeX] [PDF]

Theodorou, E. (2006). Statistical Learning of LQG controllers, Technical Report-2006-1.
[Keywords: lqg controllers, statistical learning, system identification]
[Detail] [BibTeX] [PDF]

Nakanishi, J.;Farrell, J. A.;Schaal, S. (2005). Composite adaptive control with locally weighted statistical learning, Neural Networks, 18, 1, pp.71-90.
[Keywords: adaptive control, statistical learning, composite control law, provably stable, locally weighted regression]
[Detail] [BibTeX] [PDF]

Vijayakumar, S.;D'Souza, A.;Schaal, S. (2005). Incremental online learning in high dimensions, Neural Computation, 17, 12, pp.2602-2634.
[Keywords: nonlinear regression locally weighted learning nonparametric partial least squares factor analysis confidence statistical learning dimensionality reduction principle components]
[Detail] [BibTeX] [PDF]

D'Souza, A (2004). Towards Tractable Parameter-Free Statistical Learning (Phd Thesis), Department of Computer Science, University of Southern California.
[Keywords: parameter-free statistical learning]
[Detail] [BibTeX] [PDF]

Ting, J.; D'Souza, A.; Schaal, S. (2004). Predicting EMG Activity from Neural Firing in M1 with Bayesian Backfitting, Proceedings of the 11th Joint Symposium of Neural Computation (JSNC 2004).
[Keywords: bayesian backfitting, emg prediciton, m1, variational methods, linear models, statistical learning]
[Detail] [BibTeX] [PDF]

D'Souza, A.;Vijayakumar, S.;Schaal, S. (2003). Bayesian backfitting, Proceedings of the 10th Joint Symposium on Neural Computation (JSNC 2003).
[Keywords: statistical learning, bayesian variational methods, linear regression, graphical models]
[Detail] [BibTeX] [PDF]

Ijspeert, A.;Nakanishi, J.;Schaal, S. (2003). Learning attractor landscapes for learning motor primitives, in: Becker, S.;Thrun, S.;Obermayer, K. (eds.), Advances in Neural Information Processing Systems 15, pp.1547-1554, Cambridge, MA: MIT Press.
[Keywords: learning nonlinear attractor landscapes movement primitives humanoid robotics statistical learning]
[Detail] [BibTeX] [PDF]

Schaal, S.;Atkeson, C. G.;Vijayakumar, S. (2002). Scalable techniques from nonparameteric statistics for real-time robot learning, Applied Intelligence, 17, 1, pp.49-60.
[Keywords: statistical learning, nonparametric regression, distance metric, dimensionality reduction, high dimensional learning, robot learning, real-time learning]
[Detail] [BibTeX] [PDF]

Vijayakumar, S.;D'Souza, A.;Shibata, T.;Conradt, J.;Schaal, S. (2002). Statistical learning for humanoid robots, Autonomous Robots, 12, 1, pp.59-72.
[Keywords: statistical learning, nonparametric regression, distance metric, dimensionality reduction, high dimensional learning, humanoid robotics, oculomotor control, internal models]
[Detail] [BibTeX] [PDF]

Schaal, S.;Vijayakumar, S.;D'Souza, A.;Ijspeert, A.;Nakanishi, J. (2001). Real-time statistical learning for robotics and human augmentation, in: Jarvis, R. A.;Zelinsky, A. (eds.), International Symposium on Robotics Research.
[Keywords: humanoid robotics, statistical learning, movement primitives, real-time learning]
[Detail] [BibTeX] [PDF]

Shams, L.;Schaal, S. (2001). Graph-matching vs. entropy-based methods for object detection, Neural Networks, 14, 3, pp.345-354.
[Keywords: object recognition, computer vision, statistical learning, nonparametric density estimation, graph matching, entropy, mutual information]
[Detail] [BibTeX] [PDF]

Schaal, S.;Atkeson, C. G.;Vijayakumar, S. (2000). Real-time robot learning with locally weighted statistical learning, International Conference on Robotics and Automation (ICRA2000).
[Keywords: real-time robot learning, statistical learning, humanoid robotics, finalist for overall best paper award]
[Detail] [BibTeX] [PDF]

Schaal, S. (1999). Nonparametric regression for learning nonlinear transformations, in: Ritter, H.;Cruse, H.;Dean, J. (eds.), Prerational Intelligence in Strategies, High-Level Processes and Collective Behavior, 2, pp.595-621, Kluwer Academic Publishers.
[Keywords: nonparametric regression, learning, review, statistical learning, lazy learning]
[Detail] [BibTeX] [PDF]

Schaal, S.;Atkeson, C. G. (1998). Constructive incremental learning from only local information, Neural Computation, 10, 8, pp.2047-2084.
[Keywords: statistical learning, nonparametric regression, distance metric, incremental learning, on-line learning, supersmoothing]
[Detail] [BibTeX] [PDF]

Atkeson, C. G.;Moore, A. W.;Schaal, S. (1997). Locally weighted learning, Artificial Intelligence Review, 11, 1-5, pp.11-73.
[Keywords: statistical learning, nonparametric regression, distance metric, lazy learning]
[Detail] [BibTeX] [PDF]

Atkeson, C. G.;Moore, A. W.;Schaal, S. (1997). Locally weighted learning for control, Artificial Intelligence Review, 11, 1-5, pp.75-113.
[Keywords: statistical learning, nonparametric regression, distance metric, lazy learning, learning control, reinforcement learning]
[Detail] [BibTeX] [PDF]

Vijayakumar, S.;Schaal, S. (1997). Local dimensionality reduction for locally weighted learning, International Conference on Computational Intelligence in Robotics and Automation, pp.220-225.
[Keywords: statistical learning, nonparametric regression, distance metric, dimensionality reduction, high dimensional learning]
[Detail] [BibTeX] [PDF]

Schaal, S.;Atkeson, C. G. (1996). From isolation to cooperation: An alternative of a system of experts, in: Touretzky, D. S.;Mozer, M. C.;Hasselmo, M. E. (eds.), Advances in Neural Information Processing Systems 8, pp.605-611, MIT Press.
[Keywords: statistical learning, nonparametric regression, distance metric, supersmoothing, mixture model, incremental learning]
[Detail] [BibTeX] [PDF]

Schaal, S. (1994). Nonparametric regression for learning, Conference on Adaptive Behavior and Learning, Center of Interdisciplinary Research (ZIF) Bielefeld Germany, also technical report TR-H-098 of the ATR Human Information Processing Research Laboratories.
[Keywords: nonparametric regression, review, statistical learning]
[Detail] [BibTeX] [PDF]

Atkeson, C. G.;Schaal, S. (1993). Roles for memory-based learning in robotics, Proceedings of the Sixth International Symposium on Robotics Research, pp.503-521.
[Keywords: statistical learning, robotics, nonparametric regression]
[Detail] [BibTeX]