Center for Dynamical Systems
Center for Dynamical Systems
Fanuel M.,Center for Dynamical Systems |
Alaiz C.M.,Center for Dynamical Systems |
Suykens J.A.K.,Center for Dynamical Systems
Physical Review E - Statistical, Nonlinear, and Soft Matter Physics | Year: 2017
Communities in directed networks have often been characterized as regions with a high density of links, or as sets of nodes with certain patterns of connection. Our approach for community detection combines the optimization of a quality function and a spectral clustering of a deformation of the combinatorial Laplacian, the so-called magnetic Laplacian. The eigenfunctions of the magnetic Laplacian, which we call magnetic eigenmaps, incorporate structural information. Hence, using the magnetic eigenmaps, dense communities including directed cycles can be revealed as well as "role" communities in networks with a running flow, usually discovered thanks to mixture models. Furthermore, in the spirit of the Markov stability method, an approach for studying communities at different energy levels in the network is put forward, based on a quantum mechanical system at finite temperature. © 2017 American Physical Society.
Bahari M.H.,Center for Dynamical Systems |
Bertrand A.,Center for Dynamical Systems |
Moonen M.,Center for Dynamical Systems
IEEE/ACM Transactions on Audio Speech and Language Processing | Year: 2017
Microphone arrays allow to exploit the spatial coherence between simultaneously recorded microphone signals, e.g., to perform speech enhancement, i.e., to extract a speech signal and reduce background noise. However, in systems where the microphones are not sampled in a synchronous fashion, as it is often the case in wireless acoustic sensor networks, a sampling rate offset (SRO) exists between signals recorded in different nodes, which severely affects the speech enhancement performance. To avoid this performance reduction, the SRO should be estimated and compensated for. In this paper, we propose a new approach to blind SRO estimation for an asynchronous wireless acoustic sensor network, which exploits the phase drift of the coherence between the asynchronous microphones signals. We utilize the fact that the SRO causes a linearly increasing time delay between two signals and hence a linearly increasing phase-shift in the short-time Fourier transform domain. The increasing phase shift, observed as a phase drift of the coherence between the signals, is used in a weighted least-squares framework to estimate the SRO. This method is referred to as least-squares coherence drift (LCD). Experimental results in different real-world recording and simulated scenarios show the effectiveness of LCD compared to different benchmark methods. The LCD is effective even for short signal segments. We finally demonstrate that the use of the LCD within a conventional compensation approach eliminates the performance loss due to SRO in a speech enhancement algorithm based on the multichannel Wiener filter. © 2014 IEEE.
Hunyadi B.,Center for Dynamical Systems |
Tousseyn S.,Catholic University of Leuven |
Dupont P.,Catholic University of Leuven |
Van Huffel S.,Center for Dynamical Systems |
And 2 more authors.
NeuroImage | Year: 2015
There is growing evidence for the benefits of simultaneous EEG-fMRI as a non-invasive localising tool in the presurgical evaluation of epilepsy. However, many EEG-fMRI studies fail due to the absence of interictal epileptic discharges (IEDs) on EEG. Here we present an algorithm which makes use of fMRI as sole modality to localise the epileptogenic zone (EZ). Recent studies using various model-based or data-driven fMRI analysis techniques showed that it is feasible to find activation maps which are helpful in the detection of the EZ. However, there is lack of evidence that these techniques can be used prospectively, due to (a) their low specificity, (b) selecting multiple activation maps, or (c) a widespread epileptic network indicated by the selected maps. In the current study we present a method based on independent component analysis and a cascade of classifiers that exclusively detects a single map related to interictal epileptic brain activity. In order to establish the sensitivity and specificity of the proposed method, it was evaluated on a group of 18 EEG-negative patients with a single well-defined EZ and 13 healthy controls. The results show that our method provides maps which correctly indicate the EZ in several (N= 4) EEG-negative cases but at the same time maintaining a high specificity (92%). We conclude that our fMRI-based approach can be used in a prospective manner, and can extend the applicability of fMRI to EEG-negative cases. © 2015 Elsevier Inc.
Sorensen M.,Center for Dynamical Systems |
De Lathauwer L.,Catholic University of Leuven
IEEE Transactions on Signal Processing | Year: 2017
In Part I of this paper, we have presented a link between multidimensional harmonic retrieval (MHR) and the recently proposed coupled canonical polyadic decomposition (CPD), which implies new uniqueness conditions for MHR that are more relaxed than the existing results based on a Vandermonde constrained CPD. In Part II, we explain that the coupled CPD also provides a computational framework for MHR. In particular, we present an algebraic method for MHR based on simultaneous matrix diagonalization that is guaranteed to find the exact solution in the noiseless case, under conditions discussed in Part I. Since the simultaneous matrix diagonalization method reduces the MHR problem into an eigenvalue problem, the proposed algorithm can be interpreted as an MHR generalization of the classical ESPRIT method for one-dimensional harmonic retrieval. We also demonstrate that the presented coupled CPD framework for MHR can algebraically support multirate sampling. We develop an efficient implementation which has about the same computational complexity for single-rate and multirate sampling. Numerical experiments demonstrate that by simultaneously exploiting the harmonic structure in all dimensions and making use of multirate sampling, the coupled CPD framework for MHR can lead to an improved performance compared to the conventional Vandermonde constrained CPD approaches. © 1991-2012 IEEE.
Sakai R.,Center for Dynamical Systems |
Aerts J.,Center for Dynamical Systems
BMC Proceedings | Year: 2014
Background: The sequence logo is a graphical representation of a set of aligned sequences, commonly used to depict conservation of amino acid or nucleotide sequences. Although it effectively communicates the amount of information present at every position, this visual representation falls short when the domain task is to compare between two or more sets of aligned sequences. We present a new visual presentation called a Sequence Diversity Diagram and validate our design choices with a case study. Methods: Our software was developed using the open-source program called Processing. It loads multiple sequence alignment FASTA files and a configuration file, which can be modified as needed to change the visualization. Results: The redesigned figure improves on the visual comparison of two or more sets, and it additionally encodes information on sequential position conservation. In our case study of the adenylate kinase lid domain, the Sequence Diversity Diagram reveals unexpected patterns and new insights, for example the identification of subgroups within the protein subfamily. Our future work will integrate this visual encoding into interactive visualization tools to support higher level data exploration tasks. © 2014 Sakai and Aerts; licensee BioMed Central Ltd.
Sifrim A.,Center for Dynamical Systems |
Sifrim A.,Future Health |
Popovic D.,Center for Dynamical Systems |
Popovic D.,Future Health |
And 15 more authors.
Nature Methods | Year: 2013
Massively parallel sequencing greatly facilitates the discovery of novel disease genes causing Mendelian and oligogenic disorders. However, many mutations are present in any individual genome, and identifying which ones are disease causing remains a largely open problem. We introduce eXtasy, an approach to prioritize nonsynonymous single-nucleotide variants (nSNVs) that substantially improves prediction of disease-causing variants in exome sequencing data by integrating variant impact prediction, haploinsufficiency prediction and phenotype-specific gene prioritization. © 2013 Nature America, Inc. All rights reserved.
Sorber L.,Catholic University of Leuven |
Sorber L.,Center for Dynamical Systems |
Sorber L.,Future Health |
Van Barel M.,Catholic University of Leuven |
And 3 more authors.
IEEE Journal on Selected Topics in Signal Processing | Year: 2015
We present structured data fusion (SDF) as a framework for the rapid prototyping of knowledge discovery in one or more possibly incomplete data sets. In SDF, each data set - stored as a dense, sparse, or incomplete tensor - is factorized with a matrix or tensor decomposition. Factorizations can be coupled, or fused, with each other by indicating which factors should be shared between data sets. At the same time, factors may be imposed to have any type of structure that can be constructed as an explicit function of some underlying variables. With the right choice of decomposition type and factor structure, even well-known matrix factorizations such as the eigenvalue decomposition, singular value decomposition and QR factorization can be computed with SDF. A domain specific language (DSL) for SDF is implemented as part of the software package Tensorlab, with which we offer a library of tensor decompositions and factor structures to choose from. The versatility of the SDF framework is demonstrated by means of four diverse applications, which are all solved entirely within Tensorlab's DSL. © 2015 IEEE.
Widjaja D.,Center for Dynamical Systems |
Caicedo A.,Center for Dynamical Systems |
Vlemincx E.,Catholic University of Leuven |
Van Diest I.,Catholic University of Leuven |
Van Huffel S.,Center for Dynamical Systems
PLoS ONE | Year: 2014
The variability of the heart rate (HRV) is widely studied as it contains information about the activity of the autonomic nervous system (ANS). However, HRV is influenced by breathing, independently of ANS activity. It is therefore important to include respiratory information in HRV analyses in order to correctly interpret the results. In this paper, we propose to record respiratory activity and use this information to separate the tachogram in two components: one which is related to breathing and one which contains all heart rate variations that are unrelated to respiration. Several algorithms to achieve this have been suggested in the literature, but no comparison between the methods has been performed yet. In this paper, we conduct two studies to evaluate the methods' performances to accurately decompose the tachogram in two components and to assess the robustness of the algorithms. The results show that orthogonal subspace projection and an ARMAX model yield the best performances over the two comparison studies. In addition, a real-life example of stress classification is presented to demonstrate that this approach to separate respiratory information in HRV studies can reveal changes in the heart rate variations that are otherwise masked by differing respiratory patterns. © 2014 Widjaja et al.
Laenen G.,Center for Dynamical Systems |
Ardeshirdavani A.,Center for Dynamical Systems |
Moreau Y.,Center for Dynamical Systems |
Thorrez L.,Catholic University of Leuven
Nucleic acids research | Year: 2015
Galahad (https://galahad.esat.kuleuven.be) is a web-based application for analysis of drug effects. It provides an intuitive interface to be used by anybody interested in leveraging microarray data to gain insights into the pharmacological effects of a drug, mainly identification of candidate targets, elucidation of mode of action and understanding of off-target effects. The core of Galahad is a network-based analysis method of gene expression. As an input, Galahad takes raw Affymetrix human microarray data from treatment versus control experiments and provides quality control and data exploration tools, as well as computation of differential expression. Alternatively, differential expression values can be uploaded directly. Using these differential expression values, drug target prioritization and both pathway and disease enrichment can be calculated and visualized. Drug target prioritization is based on the integration of the gene expression data with a functional protein association network. The web site is free and open to all and there is no login requirement. © The Author(s) 2015. Published by Oxford University Press on behalf of Nucleic Acids Research.
Zakeri P.,Center for Dynamical Systems |
Zakeri P.,Catholic University of Leuven |
Jeuris B.,Catholic University of Leuven |
Vandebril R.,Catholic University of Leuven |
And 2 more authors.
Bioinformatics | Year: 2014
Motivation: Various approaches based on features extracted from protein sequences and often machine learning methods have been used in the prediction of protein folds. Finding an efficient technique for integrating these different protein features has received increasing attention. In particular, kernel methods are an interesting class of techniques for integrating heterogeneous data. Various methods have been proposed to fuse multiple kernels. Most techniques for multiple kernel learning focus on learning a convex linear combination of base kernels. In addition to the limitation of linear combinations, working with such approaches could cause a loss of potentially useful information. Results: We design several techniques to combine kernel matrices by taking more involved, geometry inspired means of these matrices instead of convex linear combinations. We consider various sequence-based protein features including information extracted directly from position-specific scoring matrices and local sequence alignment. We evaluate our methods for classification on the SCOP PDB-40D benchmark dataset for protein fold recognition. The best overall accuracy on the protein fold recognition test set obtained by our methods is ∼86.7%. This is an improvement over the results of the best existing approach. Moreover, our computational model has been developed by incorporating the functional domain composition of proteins through a hybridization model. It is observed that by using our proposed hybridization model, the protein fold recognition accuracy is further improved to 89.30%. Furthermore, we investigate the performance of our approach on the protein remote homology detection problem by fusing multiple string kernels. © The Author 2014.