Header logo is


2019


no image
Convolutional neural networks: A magic bullet for gravitational-wave detection?

Gebhard, T., Kilbertus, N., Harry, I., Schölkopf, B.

Physical Review D, 100(6):063015, American Physical Society, September 2019 (article)

ei

link (url) DOI [BibTex]

2019


link (url) DOI [BibTex]


no image
Data scarcity, robustness and extreme multi-label classification

Babbar, R., Schölkopf, B.

Machine Learning, 108(8):1329-1351, September 2019, Special Issue of the ECML PKDD 2019 Journal Track (article)

ei

DOI [BibTex]

DOI [BibTex]


Series Elastic Behavior of Biarticular Muscle-Tendon Structure in a Robotic Leg
Series Elastic Behavior of Biarticular Muscle-Tendon Structure in a Robotic Leg

Ruppert, F., Badri-Spröwitz, A.

Frontiers in Neurorobotics, 64, pages: 13, 13, August 2019 (article)

dlg

Frontiers YouTube link (url) DOI [BibTex]

Frontiers YouTube link (url) DOI [BibTex]


Beyond Basins of Attraction: Quantifying Robustness of Natural Dynamics
Beyond Basins of Attraction: Quantifying Robustness of Natural Dynamics

Steve Heim, , Spröwitz, A.

IEEE Transactions on Robotics (T-RO) , 35(4), pages: 939-952, August 2019 (article)

Abstract
Properly designing a system to exhibit favorable natural dynamics can greatly simplify designing or learning the control policy. However, it is still unclear what constitutes favorable natural dynamics and how to quantify its effect. Most studies of simple walking and running models have focused on the basins of attraction of passive limit cycles and the notion of self-stability. We instead emphasize the importance of stepping beyond basins of attraction. In this paper, we show an approach based on viability theory to quantify robust sets in state-action space. These sets are valid for the family of all robust control policies, which allows us to quantify the robustness inherent to the natural dynamics before designing the control policy or specifying a control objective. We illustrate our formulation using spring-mass models, simple low-dimensional models of running systems. We then show an example application by optimizing robustness of a simulated planar monoped, using a gradient-free optimization scheme. Both case studies result in a nonlinear effective stiffness providing more robustness.

dlg

arXiv preprint arXiv:1806.08081 T-RO link (url) DOI Project Page [BibTex]

arXiv preprint arXiv:1806.08081 T-RO link (url) DOI Project Page [BibTex]


no image
Automatic Segmentation and Labelling for Robot Table Tennis Time Series

Lutz, P.

Technical University Darmstadt, Germany, August 2019 (thesis)

ei

[BibTex]

[BibTex]


Learning Variable Impedance Control for Contact Sensitive Tasks
Learning Variable Impedance Control for Contact Sensitive Tasks

Bogdanovic, M., Khadiv, M., Righetti, L.

arXiv preprint, arXiv:1907.07500, July 2019 (article)

Abstract
Reinforcement learning algorithms have shown great success in solving different problems ranging from playing video games to robotics. However, they struggle to solve delicate robotic problems, especially those involving contact interactions. Though in principle a policy outputting joint torques should be able to learn these tasks, in practice we see that they have difficulty to robustly solve the problem without any structure in the action space. In this paper, we investigate how the choice of action space can give robust performance in presence of contact uncertainties. We propose to learn a policy that outputs impedance and desired position in joint space as a function of system states without imposing any other structure to the problem. We compare the performance of this approach to torque and position control policies under different contact uncertainties. Extensive simulation results on two different systems, a hopper (floating-base) with intermittent contacts and a manipulator (fixed-base) wiping a table, show that our proposed approach outperforms policies outputting torque or position in terms of both learning rate and robustness to environment uncertainty.

mg

[BibTex]


no image
SPINDLE: End-to-end learning from EEG/EMG to extrapolate animal sleep scoring across experimental settings, labs and species

Miladinovic, D., Muheim, C., Bauer, S., Spinnler, A., Noain, D., Bandarabadi, M., Gallusser, B., Krummenacher, G., Baumann, C., Adamantidis, A., Brown, S. A., Buhmann, J. M.

PLOS Computational Biology, 15(4):1-30, Public Library of Science, April 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
A 32-channel multi-coil setup optimized for human brain shimming at 9.4T

Aghaeifar, A., Zhou, J., Heule, R., Tabibian, B., Schölkopf, B., Jia, F., Zaitsev, M., Scheffler, K.

Magnetic Resonance in Medicine, 2019, (Early View) (article)

ei

DOI [BibTex]

DOI [BibTex]


Multidimensional Contrast Limited Adaptive Histogram Equalization
Multidimensional Contrast Limited Adaptive Histogram Equalization

Stimper, V., Bauer, S., Ernstorfer, R., Schölkopf, B., Xian, R. P.

IEEE Access, 7, pages: 165437-165447, 2019 (article)

ei

arXiv link (url) DOI [BibTex]

arXiv link (url) DOI [BibTex]


no image
TD-regularized actor-critic methods

Parisi, S., Tangkaratt, V., Peters, J., Khan, M. E.

Machine Learning, 108(8):1467-1501, (Editors: Karsten Borgwardt, Po-Ling Loh, Evimaria Terzi, and Antti Ukkonen), 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
A Robustness Analysis of Inverse Optimal Control of Bipedal Walking

Rebula, J. R., Schaal, S., Finley, J., Righetti, L.

IEEE Robotics and Automation Letters, 4(4):4531-4538, 2019 (article)

mg

DOI [BibTex]

DOI [BibTex]


no image
Probabilistic solutions to ordinary differential equations as nonlinear Bayesian filtering: a new perspective

Tronarp, F., Kersting, H., Särkkä, S. H. P.

Statistics and Computing, 29(6):1297-1315, 2019 (article)

ei pn

DOI [BibTex]

DOI [BibTex]


Learning to Control Highly Accelerated Ballistic Movements on Muscular Robots
Learning to Control Highly Accelerated Ballistic Movements on Muscular Robots

Büchler, D., Calandra, R., Peters, J.

2019 (article) Submitted

Abstract
High-speed and high-acceleration movements are inherently hard to control. Applying learning to the control of such motions on anthropomorphic robot arms can improve the accuracy of the control but might damage the system. The inherent exploration of learning approaches can lead to instabilities and the robot reaching joint limits at high speeds. Having hardware that enables safe exploration of high-speed and high-acceleration movements is therefore desirable. To address this issue, we propose to use robots actuated by Pneumatic Artificial Muscles (PAMs). In this paper, we present a four degrees of freedom (DoFs) robot arm that reaches high joint angle accelerations of up to 28000 °/s^2 while avoiding dangerous joint limits thanks to the antagonistic actuation and limits on the air pressure ranges. With this robot arm, we are able to tune control parameters using Bayesian optimization directly on the hardware without additional safety considerations. The achieved tracking performance on a fast trajectory exceeds previous results on comparable PAM-driven robots. We also show that our system can be controlled well on slow trajectories with PID controllers due to careful construction considerations such as minimal bending of cables, lightweight kinematics and minimal contact between PAMs and PAMs with the links. Finally, we propose a novel technique to control the the co-contraction of antagonistic muscle pairs. Experimental results illustrate that choosing the optimal co-contraction level is vital to reach better tracking performance. Through the use of PAM-driven robots and learning, we do a small step towards the future development of robots capable of more human-like motions.

ei

Arxiv Video [BibTex]


no image
Robustifying Independent Component Analysis by Adjusting for Group-Wise Stationary Noise

Pfister*, N., Weichwald*, S., Bühlmann, P., Schölkopf, B.

Journal of Machine Learning Research, 20(147):1-50, 2019, *equal contribution (article)

ei

ArXiv Code Project page PDF link (url) Project Page Project Page [BibTex]

ArXiv Code Project page PDF link (url) Project Page Project Page [BibTex]


no image
Enhancing Human Learning via Spaced Repetition Optimization

Tabibian, B., Upadhyay, U., De, A., Zarezade, A., Schölkopf, B., Gomez Rodriguez, M.

Proceedings of the National Academy of Sciences, 116(10):3988-3993, National Academy of Sciences, 2019 (article)

ei

link (url) DOI Project Page Project Page [BibTex]

link (url) DOI Project Page Project Page [BibTex]


no image
Entropic Regularization of Markov Decision Processes

Belousov, B., Peters, J.

Entropy, 21(7):674, 2019 (article)

ei

link (url) DOI [BibTex]

link (url) DOI [BibTex]


no image
Searchers adjust their eye-movement dynamics to target characteristics in natural scenes

Rothkegel, L., Schütt, H., Trukenbrod, H., Wichmann, F. A., Engbert, R.

Scientific Reports, 9(1635), 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Spatial statistics for gaze patterns in scene viewing: Effects of repeated viewing

Trukenbrod, H. A., Barthelmé, S., Wichmann, F. A., Engbert, R.

Journal of Vision, 19(6):19, 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Rigid vs compliant contact: an experimental study on biped walking

Khadiv, M., Moosavian, S. A. A., Yousefi-Koma, A., Sadedel, M., Ehsani-Seresht, A., Mansouri, S.

Multibody System Dynamics, 45(4):379-401, 2019 (article)

mg

DOI [BibTex]

DOI [BibTex]


no image
Quantum mean embedding of probability distributions

Kübler, J. M., Muandet, K., Schölkopf, B.

Physical Review Research, 1(3):033159, American Physical Society, 2019 (article)

ei

link (url) DOI [BibTex]

link (url) DOI [BibTex]


no image
Inferring causation from time series with perspectives in Earth system sciences

Runge, J., Bathiany, S., Bollt, E., Camps-Valls, G., Coumou, D., Deyle, E., Glymour, C., Kretschmer, M., Mahecha, M., Munoz-Mari, J., van Nes, E., Peters, J., Quax, R., Reichstein, M., Scheffer, M., Schölkopf, B., Spirtes, P., Sugihara, G., Sun, J., Zhang, K., Zscheischler, J.

Nature Communications, 10(2553), 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Analysis of cause-effect inference by comparing regression errors

Blöbaum, P., Janzing, D., Washio, T., Shimizu, S., Schölkopf, B.

PeerJ Computer Science, 5, pages: e169, 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Learning Intention Aware Online Adaptation of Movement Primitives

Koert, D., Pajarinen, J., Schotschneider, A., Trick, S., Rothkopf, C., Peters, J.

IEEE Robotics and Automation Letters, 4(4):3719-3726, 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Spread-spectrum magnetic resonance imaging

Scheffler, K., Loktyushin, A., Bause, J., Aghaeifar, A., Steffen, T., Schölkopf, B.

Magnetic Resonance in Medicine, 82(3):877-885, 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
How Cognitive Models of Human Body Experience Might Push Robotics

Schürmann, T., Mohler, B. J., Peters, J., Beckerle, P.

Frontiers in Neurorobotics, 13(14), 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Dense connectomic reconstruction in layer 4 of the somatosensory cortex

Motta, A., Berning, M., Boergens, K. M., Staffler, B., Beining, M., Loomba, S., Hennig, P., Wissler, H., Helmstaedter, M.

Science, 366(6469):eaay3134, American Association for the Advancement of Science, 2019 (article)

ei pn

DOI [BibTex]

DOI [BibTex]


no image
Learning Trajectory Distributions for Assisted Teleoperation and Path Planning

Ewerton, M., Arenz, O., Maeda, G., Koert, D., Kolev, Z., Takahashi, M., Peters, J.

Frontiers in Robotics and AI, 6, pages: 89, 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Birch tar production does not prove Neanderthal behavioral complexity

Schmidt, P., Blessing, M., Rageot, M., Iovita, R., Pfleging, J., Nickel, K. G., Righetti, L., Tennie, C.

Proceedings of the National Academy of Sciences (PNAS), 116(36):17707-17711, 2019 (article)

mg

DOI [BibTex]

DOI [BibTex]


no image
Brainglance: Visualizing Group Level MRI Data at One Glance

Stelzer, J., Lacosse, E., Bause, J., Scheffler, K., Lohmann, G.

Frontiers in Neuroscience, 13(972), 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Eigendecompositions of Transfer Operators in Reproducing Kernel Hilbert Spaces

Klus, S., Schuster, I., Muandet, K.

Journal of Nonlinear Science, 2019, First Online: 21 August 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Workshops of the seventh international brain-computer interface meeting: not getting lost in translation

Huggins, J. E., Guger, C., Aarnoutse, E., Allison, B., Anderson, C. W., Bedrick, S., Besio, W., Chavarriaga, R., Collinger, J. L., Do, A. H., Herff, C., Hohmann, M., Kinsella, M., Lee, K., Lotte, F., Müller-Putz, G., Nijholt, A., Pels, E., Peters, B., Putze, F., Rupp, R. S. G., Scott, S., Tangermann, M., Tubig, P., Zander, T.

Brain-Computer Interfaces, 6(3):71-101, Taylor & Francis, 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Compatible natural gradient policy search

Pajarinen, J., Thai, H. L., Akrour, R., Peters, J., Neumann, G.

Machine Learning, 108(8):1443-1466, (Editors: Karsten Borgwardt, Po-Ling Loh, Evimaria Terzi, and Antti Ukkonen), 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Learning stable and predictive structures in kinetic systems

Pfister, N., Bauer, S., Peters, J.

Proceedings of the National Academy of Sciences (PNAS), 116(51):25405-25411, 2019 (article)

ei

DOI [BibTex]

DOI [BibTex]


no image
Fairness Constraints: A Flexible Approach for Fair Classification

Zafar, M. B., Valera, I., Gomez-Rodriguez, M., Krishna, P.

Journal of Machine Learning Research, 20(75):1-42, 2019 (article)

ei

link (url) [BibTex]

link (url) [BibTex]

2010


no image
Computationally efficient algorithms for statistical image processing: Implementation in R

Langovoy, M., Wittich, O.

(2010-053), EURANDOM, Technische Universiteit Eindhoven, December 2010 (techreport)

Abstract
In the series of our earlier papers on the subject, we proposed a novel statistical hy- pothesis testing method for detection of objects in noisy images. The method uses results from percolation theory and random graph theory. We developed algorithms that allowed to detect objects of unknown shapes in the presence of nonparametric noise of unknown level and of un- known distribution. No boundary shape constraints were imposed on the objects, only a weak bulk condition for the object's interior was required. Our algorithms have linear complexity and exponential accuracy. In the present paper, we describe an implementation of our nonparametric hypothesis testing method. We provide a program that can be used for statistical experiments in image processing. This program is written in the statistical programming language R.

ei

PDF [BibTex]

2010


PDF [BibTex]


no image
Fast Convergent Algorithms for Expectation Propagation Approximate Bayesian Inference

Seeger, M., Nickisch, H.

Max Planck Institute for Biological Cybernetics, December 2010 (techreport)

Abstract
We propose a novel algorithm to solve the expectation propagation relaxation of Bayesian inference for continuous-variable graphical models. In contrast to most previous algorithms, our method is provably convergent. By marrying convergent EP ideas from (Opper&Winther 05) with covariance decoupling techniques (Wipf&Nagarajan 08, Nickisch&Seeger 09), it runs at least an order of magnitude faster than the most commonly used EP solver.

ei

Web [BibTex]

Web [BibTex]


no image
Causal relationships between frequency bands of extracellular signals in visual cortex revealed by an information theoretic analysis

Besserve, M., Schölkopf, B., Logothetis, N., Panzeri, S.

Journal of Computational Neuroscience, 29(3):547-566, December 2010 (article)

ei

PDF DOI [BibTex]

PDF DOI [BibTex]


no image
Tackling Box-Constrained Optimization via a New Projected Quasi-Newton Approach

Kim, D., Sra, S., Dhillon, I.

SIAM Journal on Scientific Computing, 32(6):3548-3563 , December 2010 (article)

Abstract
Numerous scientific applications across a variety of fields depend on box-constrained convex optimization. Box-constrained problems therefore continue to attract research interest. We address box-constrained (strictly convex) problems by deriving two new quasi-Newton algorithms. Our algorithms are positioned between the projected-gradient [J. B. Rosen, J. SIAM, 8 (1960), pp. 181–217] and projected-Newton [D. P. Bertsekas, SIAM J. Control Optim., 20 (1982), pp. 221–246] methods. We also prove their convergence under a simple Armijo step-size rule. We provide experimental results for two particular box-constrained problems: nonnegative least squares (NNLS), and nonnegative Kullback–Leibler (NNKL) minimization. For both NNLS and NNKL our algorithms perform competitively as compared to well-established methods on medium-sized problems; for larger problems our approach frequently outperforms the competition.

ei

Web DOI [BibTex]

Web DOI [BibTex]


no image
Algorithmen zum Automatischen Erlernen von Motorfähigkeiten

Peters, J., Kober, J., Schaal, S.

at - Automatisierungstechnik, 58(12):688-694, December 2010 (article)

Abstract
Robot learning methods which allow autonomous robots to adapt to novel situations have been a long standing vision of robotics, artificial intelligence, and cognitive sciences. However, to date, learning techniques have yet to fulfill this promise as only few methods manage to scale into the high-dimensional domains of manipulator robotics, or even the new upcoming trend of humanoid robotics. If possible, scaling was usually only achieved in precisely pre-structured domains. In this paper, we investigate the ingredients for a general approach policy learning with the goal of an application to motor skill refinement in order to get one step closer towards human-like performance. For doing so, we study two major components for such an approach, i. e., firstly, we study policy learning algorithms which can be applied in the general setting of motor skill learning, and, secondly, we study a theoretically well-founded general approach to representing the required control structures for task representation and execution.

ei

Web DOI [BibTex]

Web DOI [BibTex]


no image
PAC-Bayesian Analysis of Co-clustering and Beyond

Seldin, Y., Tishby, N.

Journal of Machine Learning Research, 11, pages: 3595-3646, December 2010 (article)

ei

PDF PDF [BibTex]

PDF PDF [BibTex]


no image
Gaussian Processes for Machine Learning (GPML) Toolbox

Rasmussen, C., Nickisch, H.

Journal of Machine Learning Research, 11, pages: 3011-3015, November 2010 (article)

Abstract
The GPML toolbox provides a wide range of functionality for Gaussian process (GP) inference and prediction. GPs are specified by mean and covariance functions; we offer a library of simple mean and covariance functions and mechanisms to compose more complex ones. Several likelihood functions are supported including Gaussian and heavy-tailed for regression as well as others suitable for classification. Finally, a range of inference methods is provided, including exact and variational inference, Expectation Propagation, and Laplace's method dealing with non-Gaussian likelihoods and FITC for dealing with large regression tasks.

ei

Web [BibTex]

Web [BibTex]


no image
Cryo-EM structure and rRNA model of a translating eukaryotic 80S ribosome at 5.5-Å resolution

Armache, J-P., Jarasch, A., Anger, AM., Villa, E., Becker, T., Bhushan, S., Jossinet, F., Habeck, M., Dindar, G., Franckenberg, S., Marquez, V., Mielke, T., Thomm, M., Berninghausen, O., Beatrix, B., Söding, J., Westhof, E., Wilson, DN., Beckmann, R.

Proceedings of the National Academy of Sciences of the United States of America, 107(46):19748-19753, November 2010 (article)

Abstract
Protein biosynthesis, the translation of the genetic code into polypeptides, occurs on ribonucleoprotein particles called ribosomes. Although X-ray structures of bacterial ribosomes are available, high-resolution structures of eukaryotic 80S ribosomes are lacking. Using cryoelectron microscopy and single-particle reconstruction, we have determined the structure of a translating plant (Triticum aestivum) 80S ribosome at 5.5-Å resolution. This map, together with a 6.1-Å map of a Saccharomyces cerevisiae 80S ribosome, has enabled us to model ∼98% of the rRNA. Accurate assignment of the rRNA expansion segments (ES) and variable regions has revealed unique ES–ES and r-protein–ES interactions, providing insight into the structure and evolution of the eukaryotic ribosome.

ei

Web DOI [BibTex]

Web DOI [BibTex]


no image
Policy gradient methods

Peters, J.

Scholarpedia, 5(11):3698, November 2010 (article)

Abstract
Policy gradient methods are a type of reinforcement learning techniques that rely upon optimizing parametrized policies with respect to the expected return (long-term cumulative reward) by gradient descent. They do not suffer from many of the problems that have been marring traditional reinforcement learning approaches such as the lack of guarantees of a value function, the intractability problem resulting from uncertain state information and the complexity arising from continuous states & actions.

ei

Web DOI [BibTex]

Web DOI [BibTex]


no image
Localization of eukaryote-specific ribosomal proteins in a 5.5-Å cryo-EM map of the 80S eukaryotic ribosome

Armache, J-P., Jarasch, A., Anger, AM., Villa, E., Becker, T., Bhushan, S., Jossinet, F., Habeck, M., Dindar, G., Franckenberg, S., Marquez, V., Mielke, T., Thomm, M., Berninghausen, O., Beatrix, B., Söding, J., Westhof, E., Wilson, DN., Beckmann, R.

Proceedings of the National Academy of Sciences of the United States of America, 107(46):19754-19759, November 2010 (article)

Abstract
Protein synthesis in all living organisms occurs on ribonucleoprotein particles, called ribosomes. Despite the universality of this process, eukaryotic ribosomes are significantly larger in size than their bacterial counterparts due in part to the presence of 80 r proteins rather than 54 in bacteria. Using cryoelectron microscopy reconstructions of a translating plant (Triticum aestivum) 80S ribosome at 5.5-Å resolution, together with a 6.1-Å map of a translating Saccharomyces cerevisiae 80S ribosome, we have localized and modeled 74/80 (92.5%) of the ribosomal proteins, encompassing 12 archaeal/eukaryote-specific small subunit proteins as well as the complete complement of the ribosomal proteins of the eukaryotic large subunit. Near-complete atomic models of the 80S ribosome provide insights into the structure, function, and evolution of the eukaryotic translational apparatus.

ei

Web DOI [BibTex]

Web DOI [BibTex]


no image
Spatio-Spectral Remote Sensing Image Classification With Graph Kernels

Camps-Valls, G., Shervashidze, N., Borgwardt, K.

IEEE Geoscience and Remote Sensing Letters, 7(4):741-745, October 2010 (article)

Abstract
This letter presents a graph kernel for spatio-spectral remote sensing image classification with support vector machines (SVMs). The method considers higher order relations in the neighborhood (beyond pairwise spatial relations) to iteratively compute a kernel matrix for SVM learning. The proposed kernel is easy to compute and constitutes a powerful alternative to existing approaches. The capabilities of the method are illustrated in several multi- and hyperspectral remote sensing images acquired over both urban and agricultural areas.

ei

Web DOI [BibTex]

Web DOI [BibTex]


no image
Causal Inference Using the Algorithmic Markov Condition

Janzing, D., Schölkopf, B.

IEEE Transactions on Information Theory, 56(10):5168-5194, October 2010 (article)

Abstract
Inferring the causal structure that links $n$ observables is usually based upon detecting statistical dependences and choosing simple graphs that make the joint measure Markovian. Here we argue why causal inference is also possible when the sample size is one. We develop a theory how to generate causal graphs explaining similarities between single objects. To this end, we replace the notion of conditional stochastic independence in the causal Markov condition with the vanishing of conditional algorithmic mutual information and describe the corresponding causal inference rules. We explain why a consistent reformulation of causal inference in terms of algorithmic complexity implies a new inference principle that takes into account also the complexity of conditional probability densities, making it possible to select among Markov equivalent causal graphs. This insight provides a theoretical foundation of a heuristic principle proposed in earlier work. We also sketch some ideas on how to replace Kolmogorov complexity with decidable complexity criteria. This can be seen as an algorithmic analog of replacing the empirically undecidable question of statistical independence with practical independence tests that are based on implicit or explicit assumptions on the underlying distribution.

ei

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Recurrent Policy Gradients

Wierstra, D., Förster, A., Peters, J., Schmidhuber, J.

Logic Journal of the IGPL, 18(5):620-634, October 2010 (article)

Abstract
Reinforcement learning for partially observable Markov decision problems (POMDPs) is a challenge as it requires policies with an internal state. Traditional approaches suffer significantly from this shortcoming and usually make strong assumptions on the problem domain such as perfect system models, state-estimators and a Markovian hidden system. Recurrent neural networks (RNNs) offer a natural framework for dealing with policy learning using hidden state and require only few limiting assumptions. As they can be trained well using gradient descent, they are suited for policy gradient approaches. In this paper, we present a policy gradient method, the Recurrent Policy Gradient which constitutes a model-free reinforcement learning method. It is aimed at training limited-memory stochastic policies on problems which require long-term memories of past observations. The approach involves approximating a policy gradient for a recurrent neural network by backpropagating return-weighted characteristic eligibilities through time. Using a ‘‘Long Short-Term Memory’’ RNN architecture, we are able to outperform previous RL methods on three important benchmark tasks. Furthermore, we show that using history-dependent baselines helps reducing estimation variance significantly, thus enabling our approach to tackle more challenging, highly stochastic environments.

ei

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]