Header logo is


2017


Probabilistic Line Searches for Stochastic Optimization
Probabilistic Line Searches for Stochastic Optimization

Mahsereci, M., Hennig, P.

Journal of Machine Learning Research, 18(119):1-59, November 2017 (article)

pn

link (url) Project Page [BibTex]

2017


link (url) Project Page [BibTex]


Spinal joint compliance and actuation in a simulated bounding quadruped robot
Spinal joint compliance and actuation in a simulated bounding quadruped robot

Pouya, S., Khodabakhsh, M., Sproewitz, A., Ijspeert, A.

{Autonomous Robots}, pages: 437–452, Kluwer Academic Publishers, Springer, Dordrecht, New York, NY, Febuary 2017 (article)

dlg

link (url) DOI Project Page [BibTex]

link (url) DOI Project Page [BibTex]


Early Stopping Without a Validation Set
Early Stopping Without a Validation Set

Mahsereci, M., Balles, L., Lassner, C., Hennig, P.

arXiv preprint arXiv:1703.09580, 2017 (article)

Abstract
Early stopping is a widely used technique to prevent poor generalization performance when training an over-expressive model by means of gradient-based optimization. To find a good point to halt the optimizer, a common practice is to split the dataset into a training and a smaller validation set to obtain an ongoing estimate of the generalization performance. In this paper we propose a novel early stopping criterion which is based on fast-to-compute, local statistics of the computed gradients and entirely removes the need for a held-out validation set. Our experiments show that this is a viable approach in the setting of least-squares and logistic regression as well as neural networks.

ps pn

link (url) Project Page Project Page [BibTex]


no image
Krylov Subspace Recycling for Fast Iterative Least-Squares in Machine Learning

Roos, F. D., Hennig, P.

arXiv preprint arXiv:1706.00241, 2017 (article)

Abstract
Solving symmetric positive definite linear problems is a fundamental computational task in machine learning. The exact solution, famously, is cubicly expensive in the size of the matrix. To alleviate this problem, several linear-time approximations, such as spectral and inducing-point methods, have been suggested and are now in wide use. These are low-rank approximations that choose the low-rank space a priori and do not refine it over time. While this allows linear cost in the data-set size, it also causes a finite, uncorrected approximation error. Authors from numerical linear algebra have explored ways to iteratively refine such low-rank approximations, at a cost of a small number of matrix-vector multiplications. This idea is particularly interesting in the many situations in machine learning where one has to solve a sequence of related symmetric positive definite linear problems. From the machine learning perspective, such deflation methods can be interpreted as transfer learning of a low-rank approximation across a time-series of numerical tasks. We study the use of such methods for our field. Our empirical results show that, on regression and classification problems of intermediate size, this approach can interpolate between low computational cost and numerical precision.

pn

link (url) Project Page [BibTex]


no image
Convergence Analysis of Deterministic Kernel-Based Quadrature Rules in Misspecified Settings

Kanagawa, M., Sriperumbudur, B. K., Fukumizu, K.

Arxiv e-prints, arXiv:1709.00147v1 [math.NA], 2017 (article)

Abstract
This paper presents convergence analysis of kernel-based quadrature rules in misspecified settings, focusing on deterministic quadrature in Sobolev spaces. In particular, we deal with misspecified settings where a test integrand is less smooth than a Sobolev RKHS based on which a quadrature rule is constructed. We provide convergence guarantees based on two different assumptions on a quadrature rule: one on quadrature weights, and the other on design points. More precisely, we show that convergence rates can be derived (i) if the sum of absolute weights remains constant (or does not increase quickly), or (ii) if the minimum distance between distance design points does not decrease very quickly. As a consequence of the latter result, we derive a rate of convergence for Bayesian quadrature in misspecified settings. We reveal a condition on design points to make Bayesian quadrature robust to misspecification, and show that, under this condition, it may adaptively achieve the optimal rate of convergence in the Sobolev space of a lesser order (i.e., of the unknown smoothness of a test integrand), under a slightly stronger regularity condition on the integrand.

pn

arXiv [BibTex]

arXiv [BibTex]


no image
Fast Bayesian hyperparameter optimization on large datasets

Klein, A., Falkner, S., Bartels, S., Hennig, P., Hutter, F.

Electronic Journal of Statistics, 11, 2017 (article)

pn

[BibTex]

[BibTex]


no image
Efficiency of analytical and sampling-based uncertainty propagation in intensity-modulated proton therapy

Wahl, N., Hennig, P., Wieser, H. P., Bangert, M.

Physics in Medicine & Biology, 62(14):5790-5807, 2017 (article)

Abstract
The sensitivity of intensity-modulated proton therapy (IMPT) treatment plans to uncertainties can be quantified and mitigated with robust/min-max and stochastic/probabilistic treatment analysis and optimization techniques. Those methods usually rely on sparse random, importance, or worst-case sampling. Inevitably, this imposes a trade-off between computational speed and accuracy of the uncertainty propagation. Here, we investigate analytical probabilistic modeling (APM) as an alternative for uncertainty propagation and minimization in IMPT that does not rely on scenario sampling. APM propagates probability distributions over range and setup uncertainties via a Gaussian pencil-beam approximation into moments of the probability distributions over the resulting dose in closed form. It supports arbitrary correlation models and allows for efficient incorporation of fractionation effects regarding random and systematic errors. We evaluate the trade-off between run-time and accuracy of APM uncertainty computations on three patient datasets. Results are compared against reference computations facilitating importance and random sampling. Two approximation techniques to accelerate uncertainty propagation and minimization based on probabilistic treatment plan optimization are presented. Runtimes are measured on CPU and GPU platforms, dosimetric accuracy is quantified in comparison to a sampling-based benchmark (5000 random samples). APM accurately propagates range and setup uncertainties into dose uncertainties at competitive run-times (GPU ##IMG## [http://ej.iop.org/images/0031-9155/62/14/5790/pmbaa6ec5ieqn001.gif] {$\leqslant {5}$} min). The resulting standard deviation (expectation value) of dose show average global ##IMG## [http://ej.iop.org/images/0031-9155/62/14/5790/pmbaa6ec5ieqn002.gif] {$\gamma_{{3}\% / {3}~{\rm mm}}$} pass rates between 94.2% and 99.9% (98.4% and 100.0%). All investigated importance sampling strategies provided less accuracy at higher run-times considering only a single fraction. Considering fractionation, APM uncertainty propagation and treatment plan optimization was proven to be possible at constant time complexity, while run-times of sampling-based computations are linear in the number of fractions. Using sum sampling within APM, uncertainty propagation can only be accelerated at the cost of reduced accuracy in variance calculations. For probabilistic plan optimization, we were able to approximate the necessary pre-computations within seconds, yielding treatment plans of similar quality as gained from exact uncertainty propagation. APM is suited to enhance the trade-off between speed and accuracy in uncertainty propagation and probabilistic treatment plan optimization, especially in the context of fractionation. This brings fully-fledged APM computations within reach of clinical application.

pn

link (url) [BibTex]

link (url) [BibTex]


no image
Analytical probabilistic modeling of RBE-weighted dose for ion therapy

Wieser, H., Hennig, P., Wahl, N., Bangert, M.

Physics in Medicine and Biology (PMB), 62(23):8959-8982, 2017 (article)

pn

link (url) [BibTex]

link (url) [BibTex]


no image
Embedded interruptions and task complexity influence schema-related cognitive load progression in an abstract learning task

Wirzberger, M., Bijarsari, S. E., Rey, G. D.

Acta Psychologica, 179, pages: 30-41, Elsevier, 2017 (article)

Abstract
Cognitive processes related to schema acquisition comprise an essential source of demands in learning situations. Since the related amount of cognitive load is supposed to change over time, plausible temporal models of load progression based on different theoretical backgrounds are inspected in this study. A total of 116 student participants completed a basal symbol sequence learning task, which provided insights into underlying cognitive dynamics. Two levels of task complexity were determined by the amount of elements within the symbol sequence. In addition, interruptions due to an embedded secondary task occurred at five predefined stages over the task. Within the resulting 2x5-factorial mixed between-within design, the continuous monitoring of efficiency in learning performance enabled assumptions on relevant resource investment. From the obtained results, a nonlinear change of learning efficiency over time seems most plausible in terms of cognitive load progression. Moreover, different effects of the induced interruptions show up in conditions of task complexity, which indicate the activation of distinct cognitive mechanisms related to structural aspects of the task. Findings are discussed in the light of evidence from research on memory and information processing.

re

DOI [BibTex]

DOI [BibTex]


no image
Empirical Evidence for Resource-Rational Anchoring and Adjustment

Lieder, F., Griffiths, T. L., Huys, Q. J. M., Goodman, N. D.

Psychonomic Bulletin \& Review, 25, pages: 775-784, Springer, 2017 (article)

re

[BibTex]

[BibTex]


no image
Strategy selection as rational metareasoning

Lieder, F., Griffiths, T.

Psychological Review, 124, pages: 762-794, American Psychological Association, 2017 (article)

re

Project Page [BibTex]

Project Page [BibTex]


no image
A computerized training program for teaching people how to plan better

Lieder, F., Krueger, P. M., Callaway, F., Griffiths, T. L.

PsyArXiv, 2017 (article)

re

Project Page [BibTex]

Project Page [BibTex]


no image
Toward a rational and mechanistic account of mental effort

Shenhav, A., Musslick, S., Lieder, F., Kool, W., Griffiths, T., Cohen, J., Botvinick, M.

Annual Review of Neuroscience, 40, pages: 99-124, Annual Reviews, 2017 (article)

re

Project Page [BibTex]

Project Page [BibTex]


no image
The anchoring bias reflects rational use of cognitive resources

Lieder, F., Griffiths, T. L., Huys, Q. J. M., Goodman, N. D.

Psychonomic Bulletin \& Review, 25, pages: 762-794, Springer, 2017 (article)

re

[BibTex]

[BibTex]

2013


Quasi-Newton Methods: A New Direction
Quasi-Newton Methods: A New Direction

Hennig, P., Kiefel, M.

Journal of Machine Learning Research, 14(1):843-865, March 2013 (article)

Abstract
Four decades after their invention, quasi-Newton methods are still state of the art in unconstrained numerical optimization. Although not usually interpreted thus, these are learning algorithms that fit a local quadratic approximation to the objective function. We show that many, including the most popular, quasi-Newton methods can be interpreted as approximations of Bayesian linear regression under varying prior assumptions. This new notion elucidates some shortcomings of classical algorithms, and lights the way to a novel nonparametric quasi-Newton method, which is able to make more efficient use of available information at computational cost similar to its predecessors.

ei ps pn

website+code pdf link (url) [BibTex]

2013


website+code pdf link (url) [BibTex]


no image
Analytical probabilistic modeling for radiation therapy treatment planning

Bangert, M., Hennig, P., Oelfke, U.

Physics in Medicine and Biology, 58(16):5401-5419, 2013 (article)

ei pn

PDF DOI [BibTex]

PDF DOI [BibTex]


Motor Control Adaptation to Changes in Robot Body Dynamics for a Compliant Quadruped Robot
Motor Control Adaptation to Changes in Robot Body Dynamics for a Compliant Quadruped Robot

Pouya, S., Eckert, P., Spröwitz, A., Moc̈kel, R., Ijspeert, A. J.

In Biomimetic and Biohybrid Systems, 8064, pages: 434-437, Lecture Notes in Computer Science, Springer, Heidelberg, 2013 (incollection)

Abstract
One of the major deficiencies of current robots in comparison to living beings is the ability to adapt to new conditions either resulting from environmental changes or their own dynamics. In this work we focus on situations where the robot experiences involuntary changes in its body particularly in its limbs’ inertia. Inspired from its biological counterparts we are interested in enabling the robot to adapt its motor control to the new system dynamics. To reach this goal, we propose two different control strategies and compare their performance when handling these modifications. Our results show substantial improvements in adaptivity to body changes when the robot is aware of its new dynamics and can exploit this knowledge in synthesising new motor control.

dlg

DOI [BibTex]

DOI [BibTex]


Towards Dynamic Trot Gait Locomotion: Design, Control, and Experiments with Cheetah-cub, a Compliant Quadruped Robot
Towards Dynamic Trot Gait Locomotion: Design, Control, and Experiments with Cheetah-cub, a Compliant Quadruped Robot

Spröwitz, A., Tuleu, A., Vespignani, M., Ajallooeian, M., Badri, E., Ijspeert, A. J.

{The International Journal of Robotics Research}, 32(8):932-950, Sage Publications, Inc., Cambridge, MA, 2013 (article)

Abstract
We present the design of a novel compliant quadruped robot, called Cheetah-cub, and a series of locomotion experiments with fast trotting gaits. The robot’s leg configuration is based on a spring-loaded, pantograph mechanism with multiple segments. A dedicated open-loop locomotion controller was derived and implemented. Experiments were run in simulation and in hardware on flat terrain and with a step down, demonstrating the robot’s self-stabilizing properties. The robot reached a running trot with short flight phases with a maximum Froude number of FR = 1.30, or 6.9 body lengths per second. Morphological parameters such as the leg design also played a role. By adding distal in-series elasticity, self- stability and maximum robot speed improved. Our robot has several advantages, especially when compared with larger and stiffer quadruped robot designs. (1) It is, to the best of the authors’ knowledge, the fastest of all quadruped robots below 30 kg (in terms of Froude number and body lengths per second). (2) It shows self-stabilizing behavior over a large range of speeds with open-loop control. (3) It is lightweight, compact, and electrically powered. (4) It is cheap, easy to reproduce, robust, and safe to handle. This makes it an excellent tool for research of multi-segment legs in quadruped robots.

dlg

Youtube1 Youtube2 Youtube3 Youtube4 Youtube5 DOI Project Page [BibTex]

Youtube1 Youtube2 Youtube3 Youtube4 Youtube5 DOI Project Page [BibTex]


Horse-Like Walking, Trotting, and Galloping derived from Kinematic Motion Primitives (kMPs) and their Application to Walk/Trot Transitions in a Compliant Quadruped Robot
Horse-Like Walking, Trotting, and Galloping derived from Kinematic Motion Primitives (kMPs) and their Application to Walk/Trot Transitions in a Compliant Quadruped Robot

Moro, F., Spröwitz, A., Tuleu, A., Vespignani, M., Tsagakiris, N. G., Ijspeert, A. J., Caldwell, D. G.

Biological Cybernetics, 107(3):309-320, 2013 (article)

Abstract
This manuscript proposes a method to directly transfer the features of horse walking, trotting, and galloping to a quadruped robot, with the aim of creating a much more natural (horse-like) locomotion profile. A principal component analysis on horse joint trajectories shows that walk, trot, and gallop can be described by a set of four kinematic Motion Primitives (kMPs). These kMPs are used to generate valid, stable gaits that are tested on a compliant quadruped robot. Tests on the effects of gait frequency scaling as follows: results indicate a speed optimal walking frequency around 3.4 Hz, and an optimal trotting frequency around 4 Hz. Following, a criterion to synthesize gait transitions is proposed, and the walk/trot transitions are successfully tested on the robot. The performance of the robot when the transitions are scaled in frequency is evaluated by means of roll and pitch angle phase plots.

dlg

DOI [BibTex]

DOI [BibTex]


no image
Modelling trial-by-trial changes in the mismatch negativity

Lieder, F., Daunizeau, J., Garrido, M. I., Friston, K. J., Stephan, K. E.

{PLoS} {C}omputational {B}iology, 9(2):e1002911, Public Library of Science, 2013 (article)

re

[BibTex]

[BibTex]


no image
A neurocomputational model of the mismatch negativity

Lieder, F., Stephan, K. E., Daunizeau, J., Garrido, M. I., Friston, K. J.

{PLoS Computational Biology}, 9(11):e1003288, Public Library of Science, 2013 (article)

re

[BibTex]

[BibTex]