38 research outputs found
Proceedings of the GermEval 2021 Workshop on the Identification of Toxic, Engaging, and Fact-Claiming Comments : 17th Conference on Natural Language Processing KONVENS 2021 / FHAC at GermEval 2021: Identifying German toxic, engaging, and fact-claiming comments with ensemble learning
No full textTobias Bornheim, Niklas Grieger, Stephan Bialonsk
Massiv parallele Datenanalyse für die Erdsystemmodellierung mit dem Helmholtz Analytics Toolkit
No full textInferring complex networks from time series of dynamical systems: Pitfalls, misinterpretations, and possible solutions
Get PDFUnderstanding the dynamics of spatially extended systems represents a challenge in diverse scientific disciplines, ranging from physics and mathematics to the earth and climate sciences or the neurosciences. This challenge has stimulated the development of sophisticated data analysis approaches adopting concepts from network theory: systems are considered to be composed of subsystems (nodes) which interact with each other (represented by edges). In many studies, such complex networks of interactions have been derived from empirical time series for various spatially extended systems and have been repeatedly reported to possess the same, possibly desirable, properties (e.g. small-world characteristics and assortativity). In this thesis we study whether and how interaction networks are influenced by the analysis methodology, i.e. by the way how empirical data is acquired (the spatial and temporal sampling of the dynamics) and how nodes and edges are derived from multivariate time series. Our modeling and numerical studies are complemented by field data analyses of brain activities that unfold on various spatial and temporal scales. We demonstrate that indications of small-world characteristics and assortativity can already be expected due solely to the analysis methodology, irrespective of the actual interaction structure of the system. We develop and discuss strategies to distinguish the properties of interaction networks related to the dynamics from those spuriously induced by the analysis methodology. We show how these strategies can help to avoid misinterpretations when investigating the dynamics of spatially extended systems
CiliaCoordination
No full textIn this attached zip file, there are1. an example dataset of cilia in the zebrafish nose in Matlab format (03-02-2017_0_0_aligned.mat) and its associated metadata (03-02-2017_0_0.json)2. an example dataset of cilia of ependymal cells in the zebrafish brain in Matlab format (06-08-2022_3_8.mat_aligned.mat) and its associated metadata (06-08-2022_3_8.json)3. all Matlab functions needed, downloaded from https://github.com/Jurisch-Yaksi-lab/CiliaCoordination, corresponding to v1.0.0 on 06.02.2023published inChrista Ringers, Stephan Bialonski, Mert Ege, Anton Solovev, Jan Niklas Hansen, Inyoung Jeong, Benjamin M Friedrich, Nathalie Jurisch-Yaksi (2023) Novel analytical tools reveal that local synchronization of cilia coincides with tissue-scale metachronal waves in zebrafish multiciliated epithelia eLife 12:e77701.https://doi.org/10.7554/eLife.77701THIS DATASET IS ARCHIVED AT DANS/EASY, BUT NOT ACCESSIBLE HERE. TO VIEW A LIST OF FILES AND ACCESS THE FILES IN THIS DATASET CLICK ON THE DOI-LINK ABOV
Unraveling spurious properties of interaction networks with tailored random networks.
No full textWe investigate interaction networks that we derive from multivariate time series with methods frequently employed in diverse scientific fields such as biology, quantitative finance, physics, earth and climate sciences, and the neurosciences. Mimicking experimental situations, we generate time series with finite length and varying frequency content but from independent stochastic processes. Using the correlation coefficient and the maximum cross-correlation, we estimate interdependencies between these time series. With clustering coefficient and average shortest path length, we observe unweighted interaction networks, derived via thresholding the values of interdependence, to possess non-trivial topologies as compared to Erdös-Rényi networks, which would indicate small-world characteristics. These topologies reflect the mostly unavoidable finiteness of the data, which limits the reliability of typically used estimators of signal interdependence. We propose random networks that are tailored to the way interaction networks are derived from empirical data. Through an exemplary investigation of multichannel electroencephalographic recordings of epileptic seizures--known for their complex spatial and temporal dynamics--we show that such random networks help to distinguish network properties of interdependence structures related to seizure dynamics from those spuriously induced by the applied methods of analysis
Data-Efficient Sleep Staging with Synthetic Time Series Pretraining
Get PDFAnalyzing electroencephalographic (EEG) time series can be challenging,
especially with deep neural networks, due to the large variability among human
subjects and often small datasets. To address these challenges, various
strategies, such as self-supervised learning, have been suggested, but they
typically rely on extensive empirical datasets. Inspired by recent advances in
computer vision, we propose a pretraining task termed "frequency pretraining"
to pretrain a neural network for sleep staging by predicting the frequency
content of randomly generated synthetic time series. Our experiments
demonstrate that our method surpasses fully supervised learning in scenarios
with limited data and few subjects, and matches its performance in regimes with
many subjects. Furthermore, our results underline the relevance of frequency
information for sleep stage scoring, while also demonstrating that deep neural
networks utilize information beyond frequencies to enhance sleep staging
performance, which is consistent with previous research. We anticipate that our
approach will be advantageous across a broad spectrum of applications where EEG
data is limited or derived from a small number of subjects, including the
domain of brain-computer interfaces.Comment: 10 pages, 4 figure
Data-Efficient Sleep Staging with Synthetic Time Series Pretraining
No full textAnalyzing electroencephalographic (EEG) time series can be challenging, especially with deep neural networks, due to the large variability among human subjects and often small datasets. To address these challenges, various strategies, such as self-supervised learning, have been suggested, but they typically rely on extensive empirical datasets. Inspired by recent advances in computer vision, we propose a pretraining task termed “frequency pretraining” to pretrain a neural network for sleep staging by predicting the frequency content of randomly generated synthetic time series. Our experiments demonstrate that our method surpasses fully supervised learning in scenarios with limited data and few subjects, and matches its performance in regimes with many subjects. Furthermore, our results underline the relevance of frequency information for sleep stage scoring, while also demonstrating that deep neural networks utilize information beyond frequencies to enhance sleep staging performance, which is consistent with previous research. We anticipate that our approach will be advantageous across a broad spectrum of applications where EEG data is limited or derived from a small number of subjects, including the domain of brain-computer interfaces
