Poster
Probabilistic Recurrent State-Space Models
Andreas Doerr · Christian Daniel · Martin Schiegg · Duy Nguyen-Tuong · Stefan Schaal · Marc Toussaint · Sebastian Trimpe

Thu Jul 12th 06:15 -- 09:00 PM @ Hall B #9

State-space models (SSMs) are a highly expressive model class for learning patterns in time series data and for system identification. Deterministic versions of SSMs (e.g., LSTMs) proved extremely successful in modeling complex time series data. Fully probabilistic SSMs, however, are often found hard to train, even for smaller problems. We propose a novel model formulation and a scalable training algorithm based on doubly stochastic variational inference and Gaussian processes. This combination allows efficient incorporation of latent state temporal correlations, which we found to be key to robust training. The effectiveness of the proposed PR-SSM is evaluated on a set of real-world benchmark datasets in comparison to state-of-the-art probabilistic model learning methods. Scalability and robustness are demonstrated on a high dimensional problem.

Author Information

Andreas Doerr (Bosch Center for Artificial Intelligence, Max Planck Institute for Intelligent Systems)

https://is.tuebingen.mpg.de/person/adoerr https://www.linkedin.com/in/andreasdoerr

Christian Daniel (TU Darmstadt)
Martin Schiegg (Bosch Center for AI (BCAI))
Duy Nguyen-Tuong (Bosch Center for Artificial Intelligence)

Duy Nguyen-Tuong has been with Bosch Corporate Research in Renningen since 2011, where he works on machine learning with focus on robotics and automotive applications. Before joining Bosch Research, he studied control and automation engineering at the University of Stuttgart and the National University of Singapore. From 2007 to 2011, he was a member of the Robot Learning Laboratory at the Max Planck Institute for Biological Cybernetics, where he completed his Ph.D. studies. His main research interest is the application of machine learning techniques for model learning.

Stefan Schaal (University of Southern California)

Stefan Schaal is Professor of Computer Science, Neuroscience, and Biomedical Engineering at the University of Southern California. He is a Founding Director of the Max-Planck-Insitute for Intelligent Systems in Germany where he led the Autonomous Motion Department for several years. He was also an Invited Researcher at the ATR Computational Neuroscience Laboratory in Japan, where he held an appointment as Head of the Computational Learning Group during an international ERATO project, the Kawato Dynamic Brain Project (ERATO/JST). Before joining USC, Dr. Schaal was a postdoctoral fellow at the Department of Brain and Cognitive Sciences and the Artificial Intelligence Laboratory at MIT, an Invited Researcher at the ATR Human Information Processing Research Laboratories in Japan, and an Adjunct Assistant Professor at the Georgia Institute of Technology and at the Department of Kinesiology of the Pennsylvania State University. Dr. Schaal's research interests include topics of statistical and machine learning, neural networks, computational neuroscience, functional brain imaging, nonlinear dynamics, nonlinear control theory, and biomimetic robotics. He applies his research to problems of artificial and biological motor control and motor learning, focusing on both theoretical investigations and experiments with human subjects and anthropomorphic robot equipment. Dr. Schaal has co-authored over 400 papers in refereed journals and conferences. He is a co-founder of the "IEEE/RAS International Conference and Humanoid Robotics", and a co-founder of "Robotics Science and Systems", a highly selective new conference featuring the best work in robotics every year. Dr. Schaal served as Program Chair at these conferences and he was the Program Chair of "Simulated and Adaptive Behavior" (SAB 2004) and the "IEEE/RAS International Conference on Robotics and Automation" (ICRA 2008), the largest robotics conference in the world. Dr. Schaal is has also been an Area Chair at "Neural Information Processing Systems" (NIPS) and served as Program Committee Member of the "International Conference on Machine Learning" (ICML). Dr. Schaal serves on the editorial board of the journals "Neural Networks", "International Journal of Humanoid Robotics", and "Frontiers in Neurorobotics". Dr. Schaal is a member of the German National Academic Foundation (Studienstiftung des Deutschen Volkes), the Alexander von Humboldt Foundation, the Society For Neuroscience, the Society for Neural Control of Movement, the IEEE, and AAAS.

Marc Toussaint ((organization))
Sebastian Trimpe (Max Planck Institute for Intelligent Systems)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors