Header logo is ei


2011


no image
Statistical estimation for optimization problems on graphs

Langovoy, M., Sra, S.

In pages: 1-6, NIPS Workshop on Discrete Optimization in Machine Learning (DISCML): Uncertainty, Generalization and Feedback , December 2011 (inproceedings)

Abstract
Large graphs abound in machine learning, data mining, and several related areas. A useful step towards analyzing such graphs is that of obtaining certain summary statistics — e.g., or the expected length of a shortest path between two nodes, or the expected weight of a minimum spanning tree of the graph, etc. These statistics provide insight into the structure of a graph, and they can help predict global properties of a graph. Motivated thus, we propose to study statistical properties of structured subgraphs (of a given graph), in particular, to estimate the expected objective function value of a combinatorial optimization problem over these subgraphs. The general task is very difficult, if not unsolvable; so for concreteness we describe a more specific statistical estimation problem based on spanning trees. We hope that our position paper encourages others to also study other types of graphical structures for which one can prove nontrivial statistical estimates.

PDF Web [BibTex]

2011

PDF Web [BibTex]


no image
On the discardability of data in Support Vector Classification problems

Del Favero, S., Varagnolo, D., Dinuzzo, F., Schenato, L., Pillonetto, G.

In pages: 3210-3215, IEEE, Piscataway, NJ, USA, 50th IEEE Conference on Decision and Control and European Control Conference (CDC - ECC), December 2011 (inproceedings)

Abstract
We analyze the problem of data sets reduction for support vector classification. The work is also motivated by distributed problems, where sensors collect binary measurements at different locations moving inside an environment that needs to be divided into a collection of regions labeled in two different ways. The scope is to let each agent retain and exchange only those measurements that are mostly informative for the collective reconstruction of the decision boundary. For the case of separable classes, we provide the exact conditions and an efficient algorithm to determine if an element in the training set can become a support vector when new data arrive. The analysis is then extended to the non-separable case deriving a sufficient discardability condition and a general data selection scheme for classification. Numerical experiments relative to the distributed problem show that the proposed procedure allows the agents to exchange a small amount of the collected data to obtain a highly predictive decision boundary.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Information, learning and falsification

Balduzzi, D.

In pages: 1-4, NIPS Philosophy and Machine Learning Workshop, December 2011 (inproceedings)

Abstract
There are (at least) three approaches to quantifying information. The first, algorithmic information or Kolmogorov complexity, takes events as strings and, given a universal Turing machine, quantifies the information content of a string as the length of the shortest program producing it [1]. The second, Shannon information, takes events as belonging to ensembles and quantifies the information resulting from observing the given event in terms of the number of alternate events that have been ruled out [2]. The third, statistical learning theory, has introduced measures of capacity that control (in part) the expected risk of classifiers [3]. These capacities quantify the expectations regarding future data that learning algorithms embed into classifiers. Solomonoff and Hutter have applied algorithmic information to prove remarkable results on universal induction. Shannon information provides the mathematical foundation for communication and coding theory. However, both approaches have shortcomings. Algorithmic information is not computable, severely limiting its practical usefulness. Shannon information refers to ensembles rather than actual events: it makes no sense to compute the Shannon information of a single string – or rather, there are many answers to this question depending on how a related ensemble is constructed. Although there are asymptotic results linking algorithmic and Shannon information, it is unsatisfying that there is such a large gap – a difference in kind – between the two measures. This note describes a new method of quantifying information, effective information, that links algorithmic information to Shannon information, and also links both to capacities arising in statistical learning theory [4, 5]. After introducing the measure, we show that it provides a non-universal analog of Kolmogorov complexity. We then apply it to derive basic capacities in statistical learning theory: empirical VC-entropy and empirical Rademacher complexity. A nice byproduct of our approach is an interpretation of the explanatory power of a learning algorithm in terms of the number of hypotheses it falsifies [6], counted in two different ways for the two capacities. We also discuss how effective information relates to information gain, Shannon and mutual information.

PDF Web [BibTex]

PDF Web [BibTex]


no image
A general linear non-Gaussian state-space model: Identifiability, identification, and applications

Zhang, K., Hyvärinen, A.

In JMLR Workshop and Conference Proceedings Volume 20, pages: 113-128, (Editors: Hsu, C.-N. , W.S. Lee ), MIT Press, Cambridge, MA, USA, 3rd Asian Conference on Machine Learning (ACML), November 2011 (inproceedings)

Abstract
State-space modeling provides a powerful tool for system identification and prediction. In linear state-space models the data are usually assumed to be Gaussian and the models have certain structural constraints such that they are identifiable. In this paper we propose a non-Gaussian state-space model which does not have such constraints. We prove that this model is fully identifiable. We then propose an efficient two-step method for parameter estimation: one first extracts the subspace of the latent processes based on the temporal information of the data, and then performs multichannel blind deconvolution, making use of both the temporal information and non-Gaussianity. We conduct a series of simulations to illustrate the performance of the proposed method. Finally, we apply the proposed model and parameter estimation method on real data, including major world stock indices and magnetoencephalography (MEG) recordings. Experimental results are encouraging and show the practical usefulness of the proposed model and method.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Non-stationary correction of optical aberrations

Schuler, C., Hirsch, M., Harmeling, S., Schölkopf, B.

In pages: 659-666 , (Editors: DN Metaxas and L Quan and A Sanfeliu and LJ Van Gool), IEEE, Piscataway, NJ, USA, 13th IEEE International Conference on Computer Vision (ICCV), November 2011 (inproceedings)

Abstract
Taking a sharp photo at several megapixel resolution traditionally relies on high grade lenses. In this paper, we present an approach to alleviate image degradations caused by imperfect optics. We rely on a calibration step to encode the optical aberrations in a space-variant point spread function and obtain a corrected image by non-stationary deconvolution. By including the Bayer array in our image formation model, we can perform demosaicing as part of the deconvolution.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Learning low-rank output kernels

Dinuzzo, F., Fukumizu, K.

In JMLR Workshop and Conference Proceedings Volume 20, pages: 181-196, (Editors: Hsu, C.-N. , W.S. Lee), JMLR, Cambridge, MA, USA, 3rd Asian Conference on Machine Learning (ACML) , November 2011 (inproceedings)

Abstract
Output kernel learning techniques allow to simultaneously learn a vector-valued function and a positive semidefinite matrix which describes the relationships between the outputs. In this paper, we introduce a new formulation that imposes a low-rank constraint on the output kernel and operates directly on a factor of the kernel matrix. First, we investigate the connection between output kernel learning and a regularization problem for an architecture with two layers. Then, we show that a variety of methods such as nuclear norm regularized regression, reduced-rank regression, principal component analysis, and low rank matrix approximation can be seen as special cases of the output kernel learning framework. Finally, we introduce a block coordinate descent strategy for learning low-rank output kernels.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Spatiotemporal mapping of rhythmic activity in the inferior convexity of the macaque prefrontal cortex

Panagiotaropoulos, T., Besserve, M., Crocker, B., Kapoor, V., Tolias, A., Panzeri, S., Logothetis, N.

41(239.15), 41st Annual Meeting of the Society for Neuroscience (Neuroscience), November 2011 (poster)

Abstract
The inferior convexity of the macaque prefrontal cortex (icPFC) is known to be involved in higher order processing of sensory information mediating stimulus selection, attention and working memory. Until now, the vast majority of electrophysiological investigations of the icPFC employed single electrode recordings. As a result, relatively little is known about the spatiotemporal structure of neuronal activity in this cortical area. Here we study in detail the spatiotemporal properties of local field potentials (LFP's) in the icPFC using multi electrode recordings during anesthesia. We computed the LFP-LFP coherence as a function of frequency for thousands of pairs of simultaneously recorded sites anterior to the arcuate and inferior to the principal sulcus. We observed two distinct peaks of coherent oscillatory activity between approximately 4-10 and 15-25 Hz. We then quantified the instantaneous phase of these frequency bands using the Hilbert transform and found robust phase gradients across recording sites. The dependency of the phase on the spatial location reflects the existence of traveling waves of electrical activity in the icPFC. The dominant axis of these traveling waves roughly followed the ventral-dorsal plane. Preliminary results show that repeated visual stimulation with a 10s movie had no dramatic effect on the spatial structure of the traveling waves. Traveling waves of electrical activity in the icPFC could reflect highly organized cortical processing in this area of prefrontal cortex.

Web [BibTex]

Web [BibTex]


no image
Stability Condition for Teleoperation System with Packet Loss

Hong, A., Cho, JH., Lee, DY.

In pages: 760-761, 2011 KSME Annual Fall Conference, November 2011 (inproceedings)

Abstract
This paper focuses on the stability condition of teleoperation system where there is a packet loss in communication channel. Communication channel between master and slave cause packet loss and it obviously leads to a performance degradation and instability of teleoperation system. We consider two-channel control architecture for teleoperation system, and control inputs to remote site are produced by position of master and slave. In this paper, teleoperation system is modeled in discrete domain to include packet loss process. Also, the stability condition for teleoperation system with packet loss is discussed with input-to-state stability. Finally, the stability condition is presented in LMI approach.

[BibTex]

[BibTex]


no image
Fast removal of non-uniform camera shake

Hirsch, M., Schuler, C., Harmeling, S., Schölkopf, B.

In pages: 463-470 , (Editors: DN Metaxas and L Quan and A Sanfeliu and LJ Van Gool), IEEE, Piscataway, NJ, USA, 13th IEEE International Conference on Computer Vision (ICCV), November 2011 (inproceedings)

Abstract
Camera shake leads to non-uniform image blurs. State-of-the-art methods for removing camera shake model the blur as a linear combination of homographically transformed versions of the true image. While this is conceptually interesting, the resulting algorithms are computationally demanding. In this paper we develop a forward model based on the efficient filter flow framework, incorporating the particularities of camera shake, and show how an efficient algorithm for blur removal can be obtained. Comprehensive comparisons on a number of real-world blurry images show that our approach is not only substantially faster, but it also leads to better deblurring results.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Evaluation and Optimization of MR-Based Attenuation Correction Methods in Combined Brain PET/MR

Mantlik, F., Hofmann, M., Bezrukov, I., Schmidt, H., Kolb, A., Beyer, T., Reimold, M., Schölkopf, B., Pichler, B.

2011(MIC18.M-96), 2011 IEEE Nuclear Science Symposium, Medical Imaging Conference (NSS-MIC), October 2011 (poster)

Abstract
Combined PET/MR provides simultaneous molecular and functional information in an anatomical context with unique soft tissue contrast. However, PET/MR does not support direct derivation of attenuation maps of objects and tissues within the measured PET field-of-view. Valid attenuation maps are required for quantitative PET imaging, specifically for scientific brain studies. Therefore, several methods have been proposed for MR-based attenuation correction (MR-AC). Last year, we performed an evaluation of different MR-AC methods, including simple MR thresholding, atlas- and machine learning-based MR-AC. CT-based AC served as gold standard reference. RoIs from 2 anatomic brain atlases with different levels of detail were used for evaluation of correction accuracy. We now extend our evaluation of different MR-AC methods by using an enlarged dataset of 23 patients from the integrated BrainPET/MR (Siemens Healthcare). Further, we analyze options for improving the MR-AC performance in terms of speed and accuracy. Finally, we assess the impact of ignoring BrainPET positioning aids during the course of MR-AC. This extended study confirms the overall prediction accuracy evaluation results of the first evaluation in a larger patient population. Removing datasets affected by metal artifacts from the Atlas-Patch database helped to improve prediction accuracy, although the size of the database was reduced by one half. Significant improvement in prediction speed can be gained at a cost of only slightly reduced accuracy, while further optimizations are still possible.

Web [BibTex]

Web [BibTex]


no image
Atlas- and Pattern Recognition Based Attenuation Correction on Simultaneous Whole-Body PET/MR

Bezrukov, I., Schmidt, H., Mantlik, F., Schwenzer, N., Hofmann, M., Schölkopf, B., Pichler, B.

2011(MIC18.M-116), 2011 IEEE Nuclear Science Symposium, Medical Imaging Conference (NSS-MIC), October 2011 (poster)

Abstract
With the recent availability of clinical whole-body PET/MRI it is possible to evaluate and further develop MR-based attenuation correction methods using simultaneously acquired PET/MR data. We present first results for MRAC on patient data acquired on a fully integrated whole-body PET/MRI (Biograph mMR, Siemens) using our method that applies atlas registration and pattern recognition (ATPR) and compare them to the segmentation-based (SEG) method provided by the manufacturer. The ATPR method makes use of a database of previously aligned pairs of MR-CT volumes to predict attenuation values on a continuous scale. The robustness of the method in presence of MR artifacts was improved by location and size based detection. Lesion to liver and lesion to blood ratios (LLR and LBR) were compared for both methods on 29 iso-contour ROIs in 4 patients. ATPR showed >20% higher LBR and LLR for ROIs in and >7% near osseous tissue. For ROIs in soft tissue, both methods yielded similar ratios with max. differences <6% . For ROIs located within metal artifacts in the MR image, ATPR showed >190% higher LLR and LBR than SEG, where ratios <0.1 occured. For lesions in the neighborhood of artifacts, both ratios were >15% higher for ATPR. If artifacts in MR volumes caused by metal implants are not accounted for in the computation of attenuation maps, they can lead to a strong decrease of lesion to background ratios, even to disappearance of hot spots. Metal implants are likely to occur in the patient collective receiving combined PET/MR scans, of our first 10 patients, 3 had metal implants. Our method is currently able to account for artifacts in the pelvis caused by prostheses. The ability of the ATPR method to account for bone leads to a significant increase of LLR and LBR in osseous tissue, which supports our previous evaluations with combined PET/CT and PET/MR data. For lesions within soft tissue, lesion to background ratios of ATPR and SEG were comparable.

Web [BibTex]

Web [BibTex]


no image
Retrospective blind motion correction of MR images

Loktyushin, A., Nickisch, H., Pohmann, R.

Magnetic Resonance Materials in Physics, Biology and Medicine, 24(Supplement 1):498, 28th Annual Scientific Meeting ESMRMB, October 2011 (poster)

Abstract
We present a retrospective method, which significantly reduces ghosting and blurring artifacts due to subject motion. No modifications to the sequence (as in [2, 3]), or the use of additional equipment (as in [1]) are required. Our method iteratively searches for the transformation, that applied to the lines in k-space -- yields the sparsest Laplacian filter output in the spatial domain.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Model based reconstruction for GRE EPI

Blecher, W., Pohmann, R., Schölkopf, B., Seeger, M.

Magnetic Resonance Materials in Physics, Biology and Medicine, 24(Supplement 1):493-494, 28th Annual Scientific Meeting ESMRMB, October 2011 (poster)

Abstract
Model based nonlinear image reconstruction methods for MRI [3] are at the heart of modern reconstruction techniques (e.g.compressed sensing [6]). In general, models are expressed as a matrix equation where y and u are column vectors of k-space and image data, X model matrix and e independent noise. However, solving the corresponding linear system is not tractable. Therefore fast nonlinear algorithms that minimize a function wrt.the unknown image are the method of choice: In this work a model for gradient echo EPI, is proposed that incorporates N/2 Ghost correction and correction for field inhomogeneities. In addition to reconstruction from full data, the model allows for sparse reconstruction, joint estimation of image, field-, and relaxation-map (like [5,8] for spiral imaging), and improved N/2 ghost correction.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Attenuation correction in MR-BrainPET with segmented T1-weighted MR images of the patient’s head: A comparative study with CT

Wagenknecht, G., Rota Kops, E., Mantlik, F., Fried, E., Pilz, T., Hautzel, H., Tellmann, L., Pichler, B., Herzog, H.

In pages: 2261-2266 , IEEE, Piscataway, NJ, USA, IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC), October 2011 (inproceedings)

Abstract
Our method for attenuation correction (AC) in MR-BrainPET with segmented T1-weighted MR images of the pa-tient's head was applied to data from different MR-BrainPET scanners (Jülich, Tübingen) and compared to CT-based results. The study objectives presented in this paper are twofold. The first objective is to examine if the segmentation method developed for and successfully applied to 3D MP-RAGE data can also be used to segment other T1-weighted MR data such as 3D FLASH data. The second aim is to show if the similarity of segmented MR-based (SBA) and CT-based AC (CBA) obtained at HR+ PET can also be confirmed for BrainPET for which the new AC method is intended for. In order to reach the first objective, 14 segmented MR data sets (three 3D MP-RAGE data sets from Jülich and eleven 3D FLASH data sets from Tubingen) were compared to the resp. CT data based on the Dice coefficient and scatter plots. For bone, a CT threshold HU>;500 was applied. Dice coefficients (mean±std) for the upper cranial part of the skull, the skull above cavities, and in the caudal part including the cerebellum are 0.73±0.1, 0.79±0.04, and 0.49±0.02 for the Jülich data and 0.7U0.1, 0.72±0.1, and 0.60±0.05 for the Tubingen data. To reach the second aim, SBA and CBA were compared for six subjects based on VOI (AAL atlas) analysis. Mean absolute relative difference (maRD) values are maRD(JUFVBWl-FDG): 0.99%±0.83%, maRD(JüFVBW2-FDG): 0.90%±0.89%, and maRD(JUEP-Fluma- zenil): 1.85%±1.25% for the Jülich data and maRD(TuTP02- FDG): 2.99%±1.65%, maRD(TuNP01-FDG): 5.37%±2.29%, and maRD(TuNP02-FDG): 6.52%±1.69% for the three best-segmented Tübingen data sets. The results show similar segmentation quality for both Tl- weighted MR sequence types. The application to AC in BrainPET - hows a high similarity to CT-based AC if the standardized ACF value for bone used in SBA is in good accordance to the bone density of the patient in question.

Web DOI [BibTex]

Web DOI [BibTex]


no image
Simultaneous multimodal imaging of patients with bronchial carcinoma in a whole body MR/PET system

Brendle, C., Sauter, A., Schmidt, H., Schraml, C., Bezrukov, I., Martirosian, P., Hetzel, J., Müller, M., Claussen, C., Schwenzer, N., Pfannenberg, C.

Magnetic Resonance Materials in Physics, Biology and Medicine, 24(Supplement 1):141, 28th annual scientific meeting of the European Society for Magnetic Resonance in Medicine and Biology (ESMRB), October 2011 (poster)

Abstract
Purpose/Introduction: Lung cancer is among the most frequent cancers (1). Exact determination of tumour extent and viability is crucial for adequate therapy guidance. [18F]-FDG-PET allows accurate staging and the evaluation of therapy response based on glucose metabolism. Diffusion weighted MRI (DWI) is another promising tool for the evaluation of tumour viability (2,3). The aim of the study was the simultaneous PET-MR acquisition in lung cancer patients and correlation of PET and MR data. Subjects and Methods: Seven patients (age 38-73 years, mean 61 years) with highly suspected or known bronchial carcinoma were examined. First, a [18F]-FDG-PET/CT was performed (injected dose: 332-380 MBq). Subsequently, patients were examined at the whole-body MR/PET (Siemens Biograph mMR). The MRI is a modified 3T Verio whole body system with a magnet bore of 60 cm (max. amplitude gradients 45 mT/m, max. slew rate 200 T/m/s). Concerning the PET, the whole-body MR/PET system comprises 56 detector cassettes with a 59.4 cm transaxial and 25.8 cm axial FoV. The following parameters for PET acquisition were applied: 2 bed positions, 6 min/bed with an average uptake time of 124 min after injection (range: 110-143 min). The attenuation correction of PET data was conducted with a segmentation-based method provided by the manufacturer. Acquired PET data were reconstructed with an iterative 3D OSEM algorithm using 3 iterations and 21 subsets, Gaussian filter of 3 mm. DWI MR images were recorded simultaneously for each bed using two b-values (0/800 s/mm2). SUVmax and ADCmin were assessed in a ROI analysis. The following ratios were calculated: SUVmax(tumor)/SUVmean(liver) and ADCmin(tumor)/ADCmean(muscle). Correlation between SUV and ADC was analyzed (Pearson’s correlation). Results: Diagnostic scans could be obtained in all patients with good tumour delineation. The spatial matching of PET and DWI data was very exact. Most tumours showed a pronounced FDG-uptake in combination with decreased ADC values. Significant correlation was found between SUV and ADC ratios (r = -0.87, p = 0.0118). Discussion/Conclusion: Simultaneous MR/PET imaging of lung cancer is feasible. The whole-body MR/PET system can provide complementary information regarding tumour viability and cellularity which could facilitate a more profound tumour characterization. Further studies have to be done to evaluate the importance of these parameters for therapy decisions and monitoring

Web DOI [BibTex]

Web DOI [BibTex]


no image
Learning anticipation policies for robot table tennis

Wang, Z., Lampert, C., Mülling, K., Schölkopf, B., Peters, J.

In pages: 332-337 , (Editors: NM Amato), IEEE, Piscataway, NJ, USA, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), September 2011 (inproceedings)

Abstract
Playing table tennis is a difficult task for robots, especially due to their limitations of acceleration. A key bottleneck is the amount of time needed to reach the desired hitting position and velocity of the racket for returning the incoming ball. Here, it often does not suffice to simply extrapolate the ball's trajectory after the opponent returns it but more information is needed. Humans are able to predict the ball's trajectory based on the opponent's moves and, thus, have a considerable advantage. Hence, we propose to incorporate an anticipation system into robot table tennis players, which enables the robot to react earlier while the opponent is performing the striking movement. Based on visual observation of the opponent's racket movement, the robot can predict the aim of the opponent and adjust its movement generation accordingly. The policies for deciding how and when to react are obtained by reinforcement learning. We conduct experiments with an existing robot player to show that the learned reaction policy can significantly improve the performance of the overall system.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Estimating integrated information with TMS pulses during wakefulness, sleep and under anesthesia

Balduzzi, D.

In pages: 4717-4720 , IEEE, Piscataway, NJ, USA, 33rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (IEEE EMBC), September 2011 (inproceedings)

Abstract
This paper relates a recently proposed measure of information integration to experiments investigating the evoked high-density electroencephalography (EEG) response to transcranial magnetic stimulation (TMS) during wakefulness, early non-rapid eye movement (NREM) sleep and under anesthesia. We show that bistability, arising at the cellular and population level during NREM sleep and under anesthesia, dramatically reduces the brain’s ability to integrate information.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Improving Denoising Algorithms via a Multi-scale Meta-procedure

Burger, H., Harmeling, S.

In Pattern Recognition, pages: 206-215, (Editors: Mester, R. , M. Felsberg), Springer, Berlin, Germany, 33rd DAGM Symposium, September 2011 (inproceedings)

Abstract
Many state-of-the-art denoising algorithms focus on recovering high-frequency details in noisy images. However, images corrupted by large amounts of noise are also degraded in the lower frequencies. Thus properly handling all frequency bands allows us to better denoise in such regimes. To improve existing denoising algorithms we propose a meta-procedure that applies existing denoising algorithms across different scales and combines the resulting images into a single denoised image. With a comprehensive evaluation we show that the performance of many state-of-the-art denoising algorithms can be improved.

PDF DOI [BibTex]

PDF DOI [BibTex]


no image
Learning robot grasping from 3-D images with Markov Random Fields

Boularias, A., Kroemer, O., Peters, J.

In pages: 1548-1553 , (Editors: Amato, N.M.), IEEE, Piscataway, NJ, USA, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), September 2011 (inproceedings)

Abstract
Learning to grasp novel objects is an essential skill for robots operating in unstructured environments. We therefore propose a probabilistic approach for learning to grasp. In particular, we learn a function that predicts the success probability of grasps performed on surface points of a given object. Our approach is based on Markov Random Fields (MRF), and motivated by the fact that points that are geometrically close to each other tend to have similar grasp success probabilities. The MRF approach is successfully tested in simulation, and on a real robot using 3-D scans of various types of objects. The empirical results show a significant improvement over methods that do not utilize the smoothness assumption and classify each point separately from the others.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Neurofeedback of Fronto-Parietal Gamma-Oscillations

Grosse-Wentrup, M.

In pages: 172-175, (Editors: Müller-Putz, G.R. , R. Scherer, M. Billinger, A. Kreilinger, V. Kaiser, C. Neuper), Verlag der Technischen Universität Graz, Graz, Austria, 5th International Brain-Computer Interface Conference (BCI), September 2011 (inproceedings)

Abstract
In recent work, we have provided evidence that fronto-parietal γ-range oscillations are a cause of within-subject performance variations in brain-computer interfaces (BCIs) based on motor-imagery. Here, we explore the feasibility of using neurofeedback of fronto-parietal γ-power to induce a mental state that is beneficial for BCI-performance. We provide empirical evidence based on two healthy subjects that intentional attenuation of fronto-parietal γ-power results in an enhanced resting-state sensorimotor-rhythm (SMR). As a large resting-state amplitude of the SMR has been shown to correlate with good BCI-performance, our approach may provide a means to reduce performance variations in BCIs.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Learning inverse kinematics with structured prediction

Bocsi, B., Nguyen-Tuong, D., Csato, L., Schölkopf, B., Peters, J.

In pages: 698-703 , (Editors: NM Amato), IEEE, Piscataway, NJ, USA, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), September 2011 (inproceedings)

Abstract
Learning inverse kinematics of robots with redundant degrees of freedom (DoF) is a difficult problem in robot learning. The difficulty lies in the non-uniqueness of the inverse kinematics function. Existing methods tackle non-uniqueness by segmenting the configuration space and building a global solution from local experts. The usage of local experts implies the definition of an oracle, which governs the global consistency of the local models; the definition of this oracle is difficult. We propose an algorithm suitable to learn the inverse kinematics function in a single global model despite its multivalued nature. Inverse kinematics is approximated from examples using structured output learning methods. Unlike most of the existing methods, which estimate inverse kinematics on velocity level, we address the learning of the direct function on position level. This problem is a significantly harder. To support the proposed method, we conducted real world experiments on a tracking control task and tested our algorithms on these models.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Automatic foreground-background refocusing

Loktyushin, A., Harmeling, S.

In pages: 3445-3448, (Editors: Macq, B. , P. Schelkens), IEEE, Piscataway, NJ, USA, 18th IEEE International Conference on Image Processing (ICIP), September 2011 (inproceedings)

Abstract
A challenging problem in image restoration is to recover an image with a blurry foreground. Such images can easily occur with modern cameras, when the auto-focus aims mistakenly at the background (which will appear sharp) instead of the foreground, where usually the object of interest is. In this paper we propose an automatic procedure that (i) estimates the amount of out-of-focus blur, (ii) segments the image into foreground and background incorporating clues from the blurriness, (iii) recovers the sharp foreground, and finally (iv) blurs the background to refocus the scene. On several real photographs with blurry foreground and sharp background, we demonstrate the effectiveness and limitations of our method.

Web DOI [BibTex]

Web DOI [BibTex]


no image
Reinforcement Learning to adjust Robot Movements to New Situations

Kober, J., Oztop, E., Peters, J.

In Robotics: Science and Systems VI, pages: 33-40, (Editors: Matsuoka, Y. , H. F. Durrant-Whyte, J. Neira), MIT Press, Cambridge, MA, USA, 2010 Robotics: Science and Systems Conference (RSS), September 2011 (inproceedings)

Abstract
Many complex robot motor skills can be represented using elementary movements, and there exist efficient techniques for learning parametrized motor plans using demonstrations and self-improvement. However, in many cases, the robot currently needs to learn a new elementary movement even if a parametrized motor plan exists that covers a similar, related situation. Clearly, a method is needed that modulates the elementary movement through the meta-parameters of its representation. In this paper, we show how to learn such mappings from circumstances to meta-parameters using reinforcement learning.We introduce an appropriate reinforcement learning algorithm based on a kernelized version of the reward-weighted regression. We compare this algorithm to several previous methods on a toy example and show that it performs well in comparison to standard algorithms. Subsequently, we show two robot applications of the presented setup; i.e., the generalization of throwing movements in darts, and of hitting movements in table tennis. We show that both tasks can be learned successfully using simulated and real robots.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Simultaneous EEG Recordings with Dry and Wet Electrodes in Motor-Imagery

Saab, J., Battes, B., Grosse-Wentrup, M.

In pages: 312-315, (Editors: Müller-Putz, G.R. , R. Scherer, M. Billinger, A. Kreilinger, V. Kaiser, C. Neuper), Verlag der Technischen Universität Graz, Graz, Austria, 5th International Brain-Computer Interface Conference (BCI), September 2011 (inproceedings)

Abstract
Robust dry EEG electrodes are arguably the key to making EEG Brain-Computer Interfaces (BCIs) a practical technology. Existing studies on dry EEG electrodes can be characterized by the recording method (stand-alone dry electrodes or simultaneous recording with wet electrodes), the dry electrode technology (e.g. active or passive), the paradigm used for testing (e.g. event-related potentials), and the measure of performance (e.g. comparing dry and wet electrode frequency spectra). In this study, an active-dry electrode prototype is tested, during a motor-imagery task, with EEG-BCI in mind. It is used simultaneously with wet electrodes and assessed using classification accuracy. Our results indicate that the two types of electrodes are comparable in their performance but there are improvements to be made, particularly in finding ways to reduce motion-related artifacts.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Learning task-space tracking control with kernels

Nguyen-Tuong, D., Peters, J.

In pages: 704-709 , (Editors: Amato, N.M.), IEEE, Piscataway, NJ, USA, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), September 2011 (inproceedings)

Abstract
Task-space tracking control is essential for robot manipulation. In practice, task-space control of redundant robot systems is known to be susceptive to modeling errors. Here, data driven learning methods may present an interesting alternative approach. However, learning models for task-space tracking control from sampled data is an ill-posed problem. In particular, the same input data point can yield many different output values which can form a non-convex solution space. Because the problem is ill-posed, models cannot be learned from such data using common regression methods. While learning of task-space control mappings is globally ill-posed, it has been shown in recent work that it is locally a well-defined problem. In this paper, we use this insight to formulate a local kernel-based learning approach for online model learning for taskspace tracking control. For evaluations, we show in simulation the ability of the method for online model learning for task-space tracking control of redundant robots.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Automatic particle picking using diffusion filtering and random forest classification

Joubert, P., Nickell, S., Beck, F., Habeck, M., Hirsch, M., Schölkopf, B.

In pages: 6, International Workshop on Microscopic Image Analysis with Application in Biology (MIAAB), September 2011 (inproceedings)

Abstract
An automatic particle picking algorithm for processing electron micrographs of a large molecular complex, the 26S proteasome, is described. The algorithm makes use of a coherence enhancing diffusion filter to denoise the data, and a random forest classifier for removing false positives. It does not make use of a 3D reference model, but uses a training set of manually picked particles instead. False positive and false negative rates of around 25% to 30% are achieved on a testing set. The algorithm was developed for a specific particle, but contains steps that should be useful for developing automatic picking algorithms for other particles.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Active Versus Semi-supervised Learning Paradigm for the Classification of Remote Sensing Images

Persello, C., Bruzzone, L.

In pages: 1-15, (Editors: Bruzzone, L.), SPIE, Bellingham, WA, USA, Image and Signal Processing for Remote Sensing XVII, September 2011 (inproceedings)

Abstract
This paper presents a comparative study in order to analyze active learning (AL) and semi-supervised learning (SSL) for the classification of remote sensing (RS) images. The two learning paradigms are analyzed both from the theoretical and experimental point of view. The aim of this work is to identify the advantages and disadvantages of AL and SSL methods, and to point out the boundary conditions on the applicability of these methods with respect to both the number of available labeled samples and the reliability of classification results. In our experimental analysis, AL and SSL techniques have been applied to the classification of both synthetic and real RS data, defining different classification problems starting from different initial training sets and considering different distributions of the classes. This analysis allowed us to derive important conclusion about the use of these classification approaches and to obtain insight about which one of the two approaches is more appropriate according to the specific classification problem, the available initial training set and the available budget for the acquisition of new labeled samples.

Web DOI [BibTex]

Web DOI [BibTex]


no image
Learning elementary movements jointly with a higher level task

Kober, J., Peters, J.

In pages: 338-343 , (Editors: Amato, N.M.), IEEE, Piscataway, NJ, USA, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), September 2011 (inproceedings)

Abstract
Many motor skills consist of many lower level elementary movements that need to be sequenced in order to achieve a task. In order to learn such a task, both the primitive movements as well as the higher-level strategy need to be acquired at the same time. In contrast, most learning approaches focus either on learning to combine a fixed set of options or to learn just single options. In this paper, we discuss a new approach that allows improving the performance of lower level actions while pursuing a higher level task. The presented approach is applicable to learning a wider range motor skills, but in this paper, we employ it for learning games where the player wants to improve his performance at the individual actions of the game while still performing well at the strategy level game. We propose to learn the lower level actions using Cost-regularized Kernel Regression and the higher level actions using a form of Policy Iteration. The two approaches are coupled by their transition probabilities. We evaluate the approach on a side-stall-style throwing game both in simulation and with a real BioRob.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Adaptive nonparametric detection in cryo-electron microscopy

Langovoy, M., Habeck, M., Schölkopf, B.

In Proceedings of the 58th World Statistics Congress, pages: 4456-4461, ISI, August 2011 (inproceedings)

Abstract
We develop a novel method for detection of signals and reconstruction of images in the presence of random noise. The method uses results from percolation theory. We specifically address the problem of detection of multiple objects of unknown shapes in the case of nonparametric noise. The noise density is unknown and can be heavy-tailed. The objects of interest have unknown varying intensities. No boundary shape constraints are imposed on the objects, only a set of weak bulk conditions is required. We view the object detection problem as hypothesis testing for discrete statistical inverse problems. We present an algorithm that allows to detect greyscale objects of various shapes in noisy images. We prove results on consistency and algorithmic complexity of our procedures. Applications to cryo-electron microscopy are presented.

PDF link (url) [BibTex]

PDF link (url) [BibTex]


no image
Balancing Safety and Exploitability in Opponent Modeling

Wang, Z., Boularias, A., Mülling, K., Peters, J.

In Proceedings of the Twenty-Fifth AAAI Conference on Artificial Intelligence (AAAI 2011), pages: 1515-1520, (Editors: Burgard, W. and Roth, D.), AAAI Press, Menlo Park, CA, USA, August 2011 (inproceedings)

Abstract
Opponent modeling is a critical mechanism in repeated games. It allows a player to adapt its strategy in order to better respond to the presumed preferences of his opponents. We introduce a new modeling technique that adaptively balances exploitability and risk reduction. An opponent’s strategy is modeled with a set of possible strategies that contain the actual strategy with a high probability. The algorithm is safe as the expected payoff is above the minimax payoff with a high probability, and can exploit the opponents’ preferences when sufficient observations have been obtained. We apply them to normal-form games and stochastic games with a finite number of stages. The performance of the proposed approach is first demonstrated on repeated rock-paper-scissors games. Subsequently, the approach is evaluated in a human-robot table-tennis setting where the robot player learns to prepare to return a served ball. By modeling the human players, the robot chooses a forehand, backhand or middle preparation pose before they serve. The learned strategies can exploit the opponent’s preferences, leading to a higher rate of successful returns.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Detecting emergent processes in cellular automata with excess information

Balduzzi, D.

In Advances in Artificial Life: ECAL 2011, pages: 55-62, (Editors: Lenaerts, T. , M. Giacobini, H. Bersini, P. Bourgine, M. Dorigo, R. Doursat), MIT Press, Cambridge, MA, USA, Eleventh European Conference on the Synthesis and Simulation of Living Systems, August 2011 (inproceedings)

Abstract
Many natural processes occur over characteristic spatial and temporal scales. This paper presents tools for (i) flexibly and scalably coarse-graining cellular automata and (ii) identifying which coarse-grainings express an automaton’s dynamics well, and which express its dynamics badly. We apply the tools to investigate a range of examples in Conway’s Game of Life and Hopfield networks and demonstrate that they capture some basic intuitions about emergent processes. Finally, we formalize the notion that a process is emergent if it is better expressed at a coarser granularity.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Spatial statistics, image analysis and percolation theory

Langovoy, M., Habeck, M., Schölkopf, B.

In pages: 11, American Statistical Association, Alexandria, VA, USA, 2011 Joint Statistical Meetings (JSM), August 2011 (inproceedings)

Abstract
We develop a novel method for detection of signals and reconstruction of images in the presence of random noise. The method uses results from percolation theory. We specifically address the problem of detection of multiple objects of unknown shapes in the case of nonparametric noise. The noise density is unknown. The objects of interest have unknown varying intensities. No boundary shape constraints are imposed on the objects, only a set of weak bulk conditions is required. We view the object detection problem as a multiple hypothesis testing for discrete statistical inverse problems. We present an algorithm that allows to detect greyscale objects of various shapes in noisy images. We prove results on consistency and algorithmic complexity of our procedures. Applications to cryo-electron microscopy are presented.

PDF [BibTex]

PDF [BibTex]


no image
Two-locus association mapping in subquadratic time

Achlioptas, P., Schölkopf, B., Borgwardt, K.

In pages: 726-734, (Editors: C Apté and J Ghosh and P Smyth), ACM Press, New York, NY, USA, 17th ACM SIGKKD Conference on Knowledge Discovery and Data Mining (KDD) , August 2011 (inproceedings)

Abstract
Genome-wide association studies (GWAS) have not been able to discover strong associations between many complex human diseases and single genetic loci. Mapping these phenotypes to pairs of genetic loci is hindered by the huge number of candidates leading to enormous computational and statistical problems. In GWAS on single nucleotide polymorphisms (SNPs), one has to consider in the order of 1010 to 1014 pairs, which is infeasible in practice. In this article, we give the first algorithm for 2-locus genome-wide association studies that is subquadratic in the number, n, of SNPs. The running time of our algorithm is data-dependent, but large experiments over real genomic data suggest that it scales empirically as n3/2. As a result, our algorithm can easily cope with n ~ 107, i.e., it can efficiently search all pairs of SNPs in the human genome.

Web DOI [BibTex]

Web DOI [BibTex]


no image
A Novel Active Learning Strategy for Domain Adaptation in the Classification of Remote Sensing Images

Persello, C., Bruzzone, L.

In pages: 3720-3723 , IEEE, Piscataway, NJ, USA, IEEE International Geoscience and Remote Sensing Symposium (IGARSS), July 2011 (inproceedings)

Abstract
We present a novel technique for addressing domain adaptation problems in the classification of remote sensing images with active learning. Domain adaptation is the important problem of adapting a supervised classifier trained on a given image (source domain) to the classification of another similar (but not identical) image (target domain) acquired on a different area, or on the same area at a different time. The main idea of the proposed approach is to iteratively labeling and adding to the training set the minimum number of the most informative samples from target domain, while removing the source-domain samples that does not fit with the distributions of the classes in the target domain. In this way, the classification system exploits already available information, i.e., the labeled samples of source domain, in order to minimize the number of target domain samples to be labeled, thus reducing the cost associated to the definition of the training set for the classification of the target domain. Experimental results obtained in the classification of a hyperspectral image confirm the effectiveness of the proposed technique.

Web DOI [BibTex]

Web DOI [BibTex]


no image
Reinforcement Learning to adjust Robot Movements to New Situations

Kober, J., Oztop, E., Peters, J.

In pages: 2650-2655, (Editors: Walsh, T.), AAAI Press, Menlo Park, CA, USA, Twenty-Second International Joint Conference on Artificial Intelligence (IJCAI), July 2011 (inproceedings)

Abstract
Many complex robot motor skills can be represented using elementary movements, and there exist efficient techniques for learning parametrized motor plans using demonstrations and self-improvement. However with current techniques, in many cases, the robot currently needs to learn a new elementary movement even if a parametrized motor plan exists that covers a related situation. A method is needed that modulates the elementary movement through the meta-parameters of its representation. In this paper, we describe how to learn such mappings from circumstances to meta-parameters using reinforcement learning. In particular we use a kernelized version of the reward-weighted regression. We show two robot applications of the presented setup in robotic domains; the generalization of throwing movements in darts, and of hitting movements in table tennis. We demonstrate that both tasks can be learned successfully using simulated and real robots.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Online submodular minimization for combinatorial structures

Jegelka, S., Bilmes, J.

In pages: 345-352, (Editors: Getoor, L. , T. Scheffer), International Machine Learning Society, Madison, WI, USA, 28th International Conference on Machine Learning (ICML), July 2011 (inproceedings)

Abstract
Most results for online decision problems with structured concepts, such as trees or cuts, assume linear costs. In many settings, however, nonlinear costs are more realistic. Owing to their non-separability, these lead to much harder optimization problems. Going beyond linearity, we address online approximation algorithms for structured concepts that allow the cost to be submodular, i.e., nonseparable. In particular, we show regret bounds for three Hannan-consistent strategies that capture different settings. Our results also tighten a regret bound for unconstrained online submodular minimization.

PDF PDF Web [BibTex]

PDF PDF Web [BibTex]


no image
PAC-Bayesian Analysis of the Exploration-Exploitation Trade-off

Seldin, Y., Cesa-Bianchi, N., Laviolette, F., Auer, P., Shawe-Taylor, J., Peters, J.

In pages: 1-8, ICML Workshop on Online Trading of Exploration and Exploitation 2, July 2011 (inproceedings)

Abstract
We develop a coherent framework for integrative simultaneous analysis of the exploration-exploitation and model order selection trade-offs. We improve over our preceding results on the same subject (Seldin et al., 2011) by combining PAC-Bayesian analysis with Bernstein-type inequality for martingales. Such a combination is also of independent interest for studies of multiple simultaneously evolving martingales.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Detecting low-complexity unobserved causes

Janzing, D., Sgouritsa, E., Stegle, O., Peters, J., Schölkopf, B.

In pages: 383-391, (Editors: FG Cozman and A Pfeffer), AUAI Press, Corvallis, OR, USA, 27th Conference on Uncertainty in Artificial Intelligence (UAI), July 2011 (inproceedings)

Abstract
We describe a method that infers whether statistical dependences between two observed variables X and Y are due to a \direct" causal link or only due to a connecting causal path that contains an unobserved variable of low complexity, e.g., a binary variable. This problem is motivated by statistical genetics. Given a genetic marker that is correlated with a phenotype of interest, we want to detect whether this marker is causal or it only correlates with a causal one. Our method is based on the analysis of the location of the conditional distributions P(Y jx) in the simplex of all distributions of Y . We report encouraging results on semi-empirical data.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Support Vector Machines as Probabilistic Models

Franc, V., Zien, A., Schölkopf, B.

In Proceedings of the 28th International Conference on Machine Learning, pages: 665-672, (Editors: L Getoor and T Scheffer), International Machine Learning Society, Madison, WI, USA, ICML, July 2011 (inproceedings)

Abstract
We show how the SVM can be viewed as a maximum likelihood estimate of a class of probabilistic models. This model class can be viewed as a reparametrization of the SVM in a similar vein to the v-SVM reparametrizing the classical (C-)SVM. It is not discriminative, but has a non-uniform marginal. We illustrate the benefits of this new view by rederiving and re-investigating two established SVM-related algorithms.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Identifiability of causal graphs using functional models

Peters, J., Mooij, J., Janzing, D., Schölkopf, B.

In pages: 589-598, (Editors: FG Cozman and A Pfeffer), AUAI Press, Corvallis, OR, USA, 27th Conference on Uncertainty in Artificial Intelligence (UAI), July 2011 (inproceedings)

Abstract
This work addresses the following question: Under what assumptions on the data generating process can one infer the causal graph from the joint distribution? The approach taken by conditional independencebased causal discovery methods is based on two assumptions: the Markov condition and faithfulness. It has been shown that under these assumptions the causal graph can be identified up to Markov equivalence (some arrows remain undirected) using methods like the PC algorithm. In this work we propose an alternative by Identifiable Functional Model Classes (IFMOCs). As our main theorem we prove that if the data generating process belongs to an IFMOC, one can identify the complete causal graph. To the best of our knowledge this is the first identifiability result of this kind that is not limited to linear functional relationships. We discuss how the IFMOC assumption and the Markov and faithfulness assumptions relate to each other and explain why we believe that the IFMOC assumption can be tested more easily on given data. We further provide a practical algorithm that recovers the causal graph from finitely many data; experiments on simulated data support the theoretical fndings.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Pruning nearest neighbor cluster trees

Kpotufe, S., von Luxburg, U.

In pages: 225-232, (Editors: Getoor, L. , T. Scheffer), International Machine Learning Society, Madison, WI, USA, 28th International Conference on Machine Learning (ICML), July 2011 (inproceedings)

Abstract
Nearest neighbor ($k$-NN) graphs are widely used in machine learning and data mining applications, and our aim is to better understand what they reveal about the cluster structure of the unknown underlying distribution of points. Moreover, is it possible to identify spurious structures that might arise due to sampling variability? Our first contribution is a statistical analysis that reveals how certain subgraphs of a $k$-NN graph form a consistent estimator of the cluster tree of the underlying distribution of points. Our second and perhaps most important contribution is the following finite sample guarantee. We carefully work out the tradeoff between aggressive and conservative pruning and are able to guarantee the removal of all spurious cluster structures while at the same time guaranteeing the recovery of salient clusters. This is the first such finite sample result in the context of clustering.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Testing whether linear equations are causal: A free probability theory approach

Zscheischler, J., Janzing, D., Zhang, K.

In pages: 839-847, (Editors: Cozman, F.G. , A. Pfeffer), AUAI Press, Corvallis, OR, USA, 27th Conference on Uncertainty in Artificial Intelligence (UAI), July 2011 (inproceedings)

Abstract
We propose a method that infers whether linear relations between two high-dimensional variables X and Y are due to a causal influence from X to Y or from Y to X. The earlier proposed so-called Trace Method is extended to the regime where the dimension of the observed variables exceeds the sample size. Based on previous work, we postulate conditions that characterize a causal relation between X and Y . Moreover, we describe a statistical test and argue that both causal directions are typically rejected if there is a common cause. A full theoretical analysis is presented for the deterministic case but our approach seems to be valid for the noisy case, too, for which we additionally present an approach based on a sparsity constraint. The discussed method yields promising results for both simulated and real world data.

PDF Web [BibTex]

PDF Web [BibTex]


no image
On the information-theoretic structure of distributed measurements

Balduzzi, D.

In pages: 1-15, Elsevier Science, Amsterdam, Netherlands, 7th International Workshop on Developments of Computational Models (DCM), July 2011 (inproceedings)

Abstract
The internal structure of a measuring device, which depends on what its components are and how they are organized, determines how it categorizes its inputs. This paper presents a geometric approach to studying the internal structure of measurements performed by distributed systems such as probabilistic cellular automata. It constructs the quale, a family of sections of a suitably defined presheaf, whose elements correspond to the measurements performed by all subsystems of a distributed system. Using the quale we quantify (i) the information generated by a measurement; (ii) the extent to which a measurement is context-dependent; and (iii) whether a measurement is decomposable into independent submeasurements, which turns out to be equivalent to context-dependence. Finally, we show that only indecomposable measurements are more informative than the sum of their submeasurements.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Towards Brain-Robot Interfaces in Stroke Rehabilitation

Gomez Rodriguez, M., Grosse-Wentrup, M., Hill, J., Gharabaghi, A., Schölkopf, B., Peters, J.

In pages: 6, IEEE, Piscataway, NJ, USA, 12th International Conference on Rehabilitation Robotics (ICORR), July 2011 (inproceedings)

Abstract
A neurorehabilitation approach that combines robot-assisted active physical therapy and Brain-Computer Interfaces (BCIs) may provide an additional mileage with respect to traditional rehabilitation methods for patients with severe motor impairment due to cerebrovascular brain damage (e.g., stroke) and other neurological conditions. In this paper, we describe the design and modes of operation of a robot-based rehabilitation framework that enables artificial support of the sensorimotor feedback loop. The aim is to increase cortical plasticity by means of Hebbian-type learning rules. A BCI-based shared-control strategy is used to drive a Barret WAM 7-degree-of-freedom arm that guides a subject's arm. Experimental validation of our setup is carried out both with healthy subjects and stroke patients. We review the empirical results which we have obtained to date, and argue that they support the feasibility of future rehabilitative treatments employing this novel approach.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]


no image
Uncovering the Temporal Dynamics of Diffusion Networks

Gomez Rodriguez, M., Balduzzi, D., Schölkopf, B.

In Proceedings of the 28th International Conference on Machine Learning, pages: 561-568, (Editors: L. Getoor and T. Scheffer), Omnipress, Madison, WI, USA, ICML, July 2011 (inproceedings)

Abstract
Time plays an essential role in the diffusion of information, influence and disease over networks. In many cases we only observe when a node copies information, makes a decision or becomes infected -- but the connectivity, transmission rates between nodes and transmission sources are unknown. Inferring the underlying dynamics is of outstanding interest since it enables forecasting, influencing and retarding infections, broadly construed. To this end, we model diffusion processes as discrete networks of continuous temporal processes occurring at different rates. Given cascade data -- observed infection times of nodes -- we infer the edges of the global diffusion network and estimate the transmission rates of each edge that best explain the observed data. The optimization problem is convex. The model naturally (without heuristics) imposes sparse solutions and requires no parameter tuning. The problem decouples into a collection of independent smaller problems, thus scaling easily to networks on the order of hundreds of thousands of nodes. Experiments on real and synthetic data show that our algorithm both recovers the edges of diffusion networks and accurately estimates their transmission rates from cascade data.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Risk-Based Generalizations of f-divergences

García-García, D., von Luxburg, U., Santos-Rodríguez, R.

In pages: 417-424, (Editors: Getoor, L. , T. Scheffer), International Machine Learning Society, Madison, WI, USA, 28th International Conference on Machine Learning (ICML), July 2011 (inproceedings)

Abstract
We derive a generalized notion of f-divergences, called (f,l)-divergences. We show that this generalization enjoys many of the nice properties of f-divergences, although it is a richer family. It also provides alternative definitions of standard divergences in terms of surrogate risks. As a first practical application of this theory, we derive a new estimator for the Kulback-Leibler divergence that we use for clustering sets of vectors.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Kernel-based Conditional Independence Test and Application in Causal Discovery

Zhang, K., Peters, J., Janzing, D., Schölkopf, B.

In pages: 804-813, (Editors: FG Cozman and A Pfeffer), AUAI Press, Corvallis, OR, USA, 27th Conference on Uncertainty in Artificial Intelligence (UAI), July 2011 (inproceedings)

Abstract
Conditional independence testing is an important problem, especially in Bayesian network learning and causal discovery. Due to the curse of dimensionality, testing for conditional independence of continuous variables is particularly challenging. We propose a Kernel-based Conditional Independence test (KCI-test), by constructing an appropriate test statistic and deriving its asymptotic distribution under the null hypothesis of conditional independence. The proposed method is computationally efficient and easy to implement. Experimental results show that it outperforms other methods, especially when the conditioning set is large or the sample size is not very large, in which case other methods encounter difficulties.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Approximation Bounds for Inference using Cooperative Cut

Jegelka, S., Bilmes, J.

In pages: 577-584, (Editors: Getoor, L. , T. Scheffer), International Machine Learning Society, Madison, WI, USA, 28th International Conference on Machine Learning (ICML), July 2011 (inproceedings)

Abstract
We analyze a family of probability distributions that are characterized by an embedded combinatorial structure. This family includes models having arbitrary treewidth and arbitrary sized factors. Unlike general models with such freedom, where the “most probable explanation” (MPE) problem is inapproximable, the combinatorial structure within our model, in particular the indirect use of submodularity, leads to several MPE algorithms that all have approximation guarantees.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Multi-label cooperative cuts

Jegelka, S., Bilmes, J.

In pages: 1-4, CVPR Workshop on Inference in Graphical Models with Structured Potentials, June 2011 (inproceedings)

Abstract
Recently, a family of global, non-submodular energy functions has been proposed that is expressed as coupling edges in a graph cut. This formulation provides a rich modelling framework and also leads to efficient approximate inference algorithms. So far, the results addressed binary random variables. Here, we extend these results to the multi-label case, and combine edge coupling with move-making algorithms.

PDF Web [BibTex]

PDF Web [BibTex]


no image
Submodularity beyond submodular energies: coupling edges in graph cuts

Jegelka, S., Bilmes, J.

In pages: 1897-1904, IEEE, Piscataway, NJ, USA, IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2011 (inproceedings)

Abstract
We propose a new family of non-submodular global energy functions that still use submodularity internally to couple edges in a graph cut. We show it is possible to develop an efficient approximation algorithm that, thanks to the internal submodularity, can use standard graph cuts as a subroutine. We demonstrate the advantages of edge coupling in a natural setting, namely image segmentation. In particular, for finestructured objects and objects with shading variation, our structured edge coupling leads to significant improvements over standard approaches.

PDF Web DOI [BibTex]

PDF Web DOI [BibTex]