30 results on '"Köster, Felix"'
Search Results
2. Data-Driven Acceleration of Multi-Physics Simulations
- Author
-
Meinecke, Stefan, Selig, Malte, Köster, Felix, Knorr, Andreas, and Lüdge, Kathy
- Subjects
Physics - Computational Physics - Abstract
Multi-physics simulations play a crucial role in understanding complex systems. However, their computational demands are often prohibitive due to high dimensionality and complex interactions, such that actual calculations often rely on approximations. To address this, we introduce a data-driven approach to approximate interactions among degrees of freedom of no direct interest and thus significantly reduce computational costs. Focusing on a semiconductor laser as a case study, we demonstrate the superiority of this method over traditional analytical approximations in both accuracy and efficiency. Our approach streamlines simulations, offering promise for complex multi-physics systems, especially for scenarios requiring a large number of individual simulations., Comment: The simulation code and the regression code is available on GitHub under MIT license (https://github.com/stmeinecke/derrom)
- Published
- 2024
3. Data-Driven Forecasting of Non-Equilibrium Solid-State Dynamics
- Author
-
Meinecke, Stefan, Köster, Felix, Christiansen, Dominik, Lüdge, Kathy, Knorr, Andreas, and Selig, Malte
- Subjects
Physics - Computational Physics - Abstract
We present a data-driven approach to efficiently approximate nonlinear transient dynamics in solid-state systems. Our proposed machine-learning model combines a dimensionality reduction stage with a nonlinear vector autoregression scheme. We report an outstanding time-series forecasting performance combined with an easy to deploy model and an inexpensive training routine. Our results are of great relevance as they have the potential to massively accelerate multi-physics simulation software and thereby guide to future development of solid-state based technologies., Comment: The simulation code and the regression code is available on GitHub under MIT license (https://github.com/stmeinecke/derrom)
- Published
- 2024
- Full Text
- View/download PDF
4. Attention-Enhanced Reservoir Computing
- Author
-
Köster, Felix, Kanno, Kazutaka, Ohkubo, Jun, and Uchida, Atsushi
- Subjects
Computer Science - Emerging Technologies ,Computer Science - Machine Learning - Abstract
Photonic reservoir computing has been successfully utilized in time-series prediction as the need for hardware implementations has increased. Prediction of chaotic time series remains a significant challenge, an area where the conventional reservoir computing framework encounters limitations of prediction accuracy. We introduce an attention mechanism to the reservoir computing model in the output stage. This attention layer is designed to prioritize distinct features and temporal sequences, thereby substantially enhancing the prediction accuracy. Our results show that a photonic reservoir computer enhanced with the attention mechanism exhibits improved prediction capabilities for smaller reservoirs. These advancements highlight the transformative possibilities of reservoir computing for practical applications where accurate prediction of chaotic time series is crucial.
- Published
- 2023
5. The role of delay-times in delay-based Photonic Reservoir Computing
- Author
-
Hülser, Tobias, Köster, Felix, Jaurigue, Lina, and Lüdge, Kathy
- Subjects
Physics - Computational Physics ,Computer Science - Emerging Technologies - Abstract
Delay-based reservoir computing has gained a lot of attention due to the relative simplicity with which this concept can be implemented in hardware. However,there is still an misconception about the relationship between the delay-time and the input clock-cycle which has noticeable consequences for the performance. We review the existing literature on this subject and introduce the concept of delay-based reservoir computing in a manner that demonstrates that there is no predefined relationship between these two times-scales. Further, we discuss ways to improve the computing performance of a reservoir formed by delay-coupled oscillators and show the crucial impact of delay-time tuning in those multi-delay systems., Comment: 16 pages, 10 figures, invited review
- Published
- 2021
6. Limitations of the Recall Capabilities in Delay-Based Reservoir Computing Systems
- Author
-
Köster, Felix, Ehlert, Dominik, and Lüdge, Kathy
- Published
- 2023
- Full Text
- View/download PDF
7. Master memory function for delay-based reservoir computers with single-variable dynamics
- Author
-
Köster, Felix, Yanchuk, Serhiy, and Lüdge, Kathy
- Subjects
Computer Science - Emerging Technologies ,Computer Science - Machine Learning ,Nonlinear Sciences - Adaptation and Self-Organizing Systems - Abstract
We show that many delay-based reservoir computers considered in the literature can be characterized by a universal master memory function (MMF). Once computed for two independent parameters, this function provides linear memory capacity for any delay-based single-variable reservoir with small inputs. Moreover, we propose an analytical description of the MMF that enables its efficient and fast computation. Our approach can be applied not only to reservoirs governed by known dynamical rules such as Mackey-Glass or Ikeda-like systems but also to reservoirs whose dynamical model is not available. We also present results comparing the performance of the reservoir computer and the memory capacity given by the MMF., Comment: To be published
- Published
- 2021
8. Limitations of the recall capabilities in delay based reservoir computing systems
- Author
-
Köster, Felix, Ehlert, Dominik, and Lüdge, Kathy
- Subjects
Computer Science - Emerging Technologies ,Computer Science - Machine Learning ,Nonlinear Sciences - Adaptation and Self-Organizing Systems ,Nonlinear Sciences - Chaotic Dynamics ,Physics - Optics - Abstract
We analyze the memory capacity of a delay based reservoir computer with a Hopf normal form as nonlinearity and numerically compute the linear as well as the higher order recall capabilities. A possible physical realisation could be a laser with external cavity, for which the information is fed via electrical injection. A task independent quantification of the computational capability of the reservoir system is done via a complete orthonormal set of basis functions. Our results suggest that even for constant readout dimension the total memory capacity is dependent on the ratio between the information input period, also called the clock cycle, and the time delay in the system. Optimal performance is found for a time delay about 1.6 times the clock cycle
- Published
- 2020
- Full Text
- View/download PDF
9. Insight into Delay Based Reservoir Computing via Eigenvalue Analysis
- Author
-
Köster, Felix, Yanchuk, Serhiy, and Lüdge, Kathy
- Subjects
Computer Science - Machine Learning ,Mathematics - Dynamical Systems ,Nonlinear Sciences - Adaptation and Self-Organizing Systems ,Statistics - Machine Learning - Abstract
In this paper we give a profound insight into the computation capability of delay-based reservoir computing via an eigenvalue analysis. We concentrate on the task-independent memory capacity to quantify the reservoir performance and compare these with the eigenvalue spectrum of the dynamical system. We show that these two quantities are deeply connected, and thus the reservoir computing performance is predictable by analyzing the small signal response of the reservoir. Our results suggest that any dynamical system used as a reservoir can be analyzed in this way. We apply our method exemplarily to a photonic laser system with feedback and compare the numerically computed recall capabilities with the eigenvalue spectrum. Optimal performance is found for a system with the eigenvalues having real parts close to zero and off-resonant imaginary parts., Comment: New Journal Submission
- Published
- 2020
- Full Text
- View/download PDF
10. Deep Time-Delay Reservoir Computing: Dynamics and Memory Capacity
- Author
-
Goldmann, Mirko, Köster, Felix, Lüdge, Kathy, and Yanchuk, Serhiy
- Subjects
Nonlinear Sciences - Adaptation and Self-Organizing Systems ,Computer Science - Machine Learning ,Mathematics - Dynamical Systems - Abstract
The Deep Time-Delay Reservoir Computing concept utilizes unidirectionally connected systems with time-delays for supervised learning. We present how the dynamical properties of a deep Ikeda-based reservoir are related to its memory capacity (MC) and how that can be used for optimization. In particular, we analyze bifurcations of the corresponding autonomous system and compute conditional Lyapunov exponents, which measure the generalized synchronization between the input and the layer dynamics. We show how the MC is related to the systems distance to bifurcations or magnitude of the conditional Lyapunov exponent. The interplay of different dynamical regimes leads to a adjustable distribution between linear and nonlinear MC. Furthermore, numerical simulations show resonances between clock cycle and delays of the layers in all degrees of the MC. Contrary to MC losses in a single-layer reservoirs, these resonances can boost separate degrees of the MC and can be used, e.g., to design a system with maximum linear MC. Accordingly, we present two configurations that empower either high nonlinear MC or long time linear MC.
- Published
- 2020
11. Deriving task specific performance from the information processing capacity of a reservoir computer
- Author
-
Hülser Tobias, Köster Felix, Lüdge Kathy, and Jaurigue Lina
- Subjects
information processing capacity ,memory capacity ,nonlinear oscillator ,reservoir computing ,Physics ,QC1-999 - Abstract
In the reservoir computing literature, the information processing capacity is frequently used to characterize the computing capabilities of a reservoir. However, it remains unclear how the information processing capacity connects to the performance on specific tasks. We demonstrate on a set of standard benchmark tasks that the total information processing capacity correlates poorly with task specific performance. Further, we derive an expression for the normalized mean square error of a task as a weighted function of the individual information processing capacities. Mathematically, the derivation requires the task to have the same input distribution as used to calculate the information processing capacities. We test our method on a range of tasks that violate this requirement and find good qualitative agreement between the predicted and the actual errors as long as the task input sequences do not have long autocorrelation times. Our method offers deeper insight into the principles governing reservoir computing performance. It also increases the utility of the evaluation of information processing capacities, which are typically defined on i.i.d. input, even if specific tasks deliver inputs stemming from different distributions. Moreover, it offers the possibility of reducing the experimental cost of optimizing physical reservoirs, such as those implemented in photonic systems.
- Published
- 2022
- Full Text
- View/download PDF
12. Screening auf psychische Komorbiditäten in der Dermatologie: Erfolgreiche Implementierung eines Screenings auf psychische Komorbiditäten im Bereich der stationären dermatologischen Versorgung
- Author
-
Köster, Felix-Wilhelm, Kohlmann, Sebastian, Loeper, Siobhan, Löwe, Bernd, and Schneider, Stefan W.
- Published
- 2021
- Full Text
- View/download PDF
13. Anticipating emission-sensitive traffic management strategies for dynamic delivery routing
- Author
-
Köster, Felix, Ulmer, Marlin W., Mattfeld, Dirk C., and Hasle, Geir
- Published
- 2018
- Full Text
- View/download PDF
14. Budgeting Time for Dynamic Vehicle Routing with Stochastic Customer Requests
- Author
-
Ulmer, Marlin W., Mattfeld, Dirk C., and Köster, Felix
- Published
- 2018
15. Dynamic Routing: Anticipation of Emission-Sensitive Traffic Management
- Author
-
Köster, Felix, Ulmer, Marlin W., and Mattfeld, Dirk C.
- Published
- 2017
- Full Text
- View/download PDF
16. Data-driven forecasting of nonequilibrium solid-state dynamics
- Author
-
Meinecke, Stefan, primary, Köster, Felix, additional, Christiansen, Dominik, additional, Lüdge, Kathy, additional, Knorr, Andreas, additional, and Selig, Malte, additional
- Published
- 2023
- Full Text
- View/download PDF
17. Time series forecasting with delay-based reservoir computing: Analysis and optimization
- Author
-
Köster, Felix
- Subjects
500 Naturwissenschaften und Mathematik::510 Mathematik::519 Wahrscheinlichkeiten, angewandte Mathematik ,nonlinear dynamics ,500 Naturwissenschaften und Mathematik::510 Mathematik::518 Numerische Analysis ,machine learning ,500 Naturwissenschaften und Mathematik::530 Physik::535 Licht, Infrarot- und Ultraviolettphänomene ,laserphysics ,time series forecasting ,reservoir computing - Abstract
This thesis presents a comprehensive analysis of delay-based reservoir computation. Chapter 1 provides a brief motivation for the importance of studying dynamical systems in various research areas and the importance of being able to make predictions for such systems. It is emphasized that physics is reaching its limits in terms of the use of modern computers in predicting the most complex dynamical systems known. This is especially due to the physical limitations of the hardware implementations themselves, which in recent years have increasingly diminished the progress of computing power. Therefore, a paradigm shift is motivated, which describes the use of dynamical systems themselves as computational units for the prediction of other dynamical systems, opening new paths for hardware implementations. This concept is called "reservoir computing". A special case of this hardware implementation, called delay-based reservoir computing, is motivated and analyzed, which is based on dynamical systems with a time-delayed feedback loop. Chapter 2 provides a detailed introduction to theories and mathematical methods used in this thesis. The concepts of dynamical systems, eigenvalue analysis, machine learning, and delay-based reservoir computation are explained in detail. A strong focus is placed on dynamical systems with a time-delayed feedback signal. Furthermore, the introduction of linear and nonlinear memory capacity is described in detail, which enables a task-independent quantification of the computational power of a reservoir computer. In addition, all algorithms are presented and explained, including computing fixed points and eigenvalues, as well as methods for simulating all time series used in this thesis. In chapter 3, the linear and nonlinear memory capacity is used to show the fundamental limitations of a delay-based reservoir computer by finding memory gaps related to long time delays τ . In addition, degraded performance is identified at resonances between the delay time τ and the clock cycle T . As a result, it is found that the delay time τ cannot be increased indefinitely without either the emergence of the above-mentioned recall gaps or the need to increase the clock cycle T . Through this analysis, an optimal delay time of τ = √2T is identified as a starting point without prior knowledge of the training and target data. It should be noted that these results may need to be adjusted if additional information about the actual data set is obtained. Chapter 4 provides further insight into the inner workings of the delay-based reservoir calculation through an eigenvalue analysis. This analysis reinforces the results from Chapter 3 through a mathematical linearization approach. The imaginary parts of the eigenvalues indicate that resonances should be avoided because they reduce the usable phase space for the reservoir computer, while the real parts show that operating points close to criticality are most favorable for tasks requiring long-term memory, while operating points away from criticality yield faster, more nonlinear systems with less memory. Chapter 5 presents an analytic formula for the linear memory capacity of an arbitrary one-dimensional delay-based reservoir computer, referred to as the "master memory function." This is an important result of this work because it allows the prediction of the linear memory capacity of delay-based reservoir computers for systems with unknown models using measurements. The "master memory function" also exhibits a universal property. This property indicates that all delay-based reservoir computers driven by a small input signal, such that their respective linearizations yield the same system, also have the same linear memory capacity. The possibility of extending the formula to multiple time-delayed feedbacks and higher dimensionality is discussed, and further exploration is recommended for interested readers. Chapter 6 compares delay-based reservoir computers with other widely used approaches to time series prediction, in particular the nonlinear autoregressive model and the kernel trick in the reproducing kernel Hilbert space. It is shown that reservoir computers are able to make predictions of comparable quality on tasks with particularly long correlations, due to the inherent memory of reservoir-based approaches. In the final chapter 7, a hybrid approach is presented to improve each aspect of the reservoir computing scheme. A data-driven forecasting method called SINDy is applied to the same dataset to derive dynamic models. These models are used in conjunction with a delay-based reservoir computer, resulting in significant improvements in shortterm forecast accuracy and long-term forecast stability, while reducing the complexity of the required models. In summary, this work provides insight into the inner workings of delay-based reservoir computers, assists in determining optimal operating points, and provides an analytical description of the linear memory capacity. The delay-based reservoir computer is placed within the broader framework of data-driven time series forecasting. This thesis also introduces the concept of a hybrid data-driven approach to the reservoir computing community, with promising results., Diese Arbeit stellt eine umfassende Analyse der verzögerungsbasierten Reservoirberechnung dar. Kapitel 1 liefert eine kurze Motivation für die Bedeutung des Studiums dynamischer Systeme in verschiedenen Forschungsbereichen und betont die Wichtigkeit, Vorhersagen für dynamische Systeme machen zu können. Dabei wird hervorgehoben, dass die Physik im Hinblick der Nutzung des modernen Computers bei der Vorhersage der komplexesten bekannten dynamischen Systeme an ihre Grenzen stößt. Dies liegt insbesondere an den physikalischen Limitationen der Hardware-Implementierungen selbst, welche in den letzten Jahren immer mehr den Fortschritt von Rechenleistung vermindert haben. Daher wird ein Paradigmenwechsel motiviert, der die Nutzung von dynamischen Systemen selbst als Recheneinheiten zur Vorhersage anderer dynamischer Systeme beschreibt. Diese Konzepts nennt sich „Reservoir-Computing", also das Berechnen mithilfe eines Reservoirs gegeben durch ein dynamisches System. Ein spezieller Fall dieser Hardware-Implementierung, genannt verzögerungsbasierte Reservoirberechnung, wird analysiert, welches auf dynamischen Systemen mit einer zeitverzögerten Rückkopplung basiert. Kapitel 2 dieser Dissertation, bietet eine detaillierte Einführung über alle in dieser Arbeit verwendeten Theorien und mathematischen Methoden. Die Konzepte der dynamischen Systeme, der Eigenwertanalyse, des maschinellen Lernens und der verzögerungsbasierten Reservoirberechnung werden im Detail erläutert. Ein starker Fokus wird auf dynamische System mit einer zeitverzögerten Rückkopplung gelegt. Außerdem wird im Detail die Einführung der linearen und nichtlinearen Speicherkapazität beschrieben, welche eine aufgabenunabhängige Quantifizierung der Rechenleistung eines Reservoir-Computers ergibt. Darüber hinaus werden alle Algorithmen und rechenintensiven Methoden vorgestellt und eingeführt, einschließlich numerischer Algorithmen zur Berechnung von Fixpunkten und Eigenwerten, sowie Methoden zur Simulation aller in dieser Arbeit verwendeten Zeitserien. In Kapitel 3 wird die lineare und nichtlineare Speicherkapazität genutzt, um die grundlegenden Grenzen eines verzögerungsbasierte Reservoir-Computers aufzuzeigen, in dem Erinnerungslücken in Bezug auf lange Zeitverzögerungen τ gefunden werden. Außerdem wird eine verminderte Leistung bei Resonanzen zwischen der Verzögerungszeit τ und dem Taktzyklus T identifiziert. Als Ergebnis wird festgestellt, dass die Verzögerungszeit τ für die Rückkopplung nicht unbegrenzt erhöht werden kann, ohne dass entweder die oben erwähnten Abruflücken auftreten oder der Taktzyklus T erhöht werden muss. Durch diese Analyse wird eine optimale Verzögerungszeit von τ = √2T als Ausgangspunkt ohne vorherige Kenntnis der Trainings- und Zieldaten identifiziert. Es sei darauf hingewiesen, dass diese Ergebnisse möglicherweise angepasst werden müssen, wenn zusätzliche Informationen über den tatsächlichen Datensatz eingeholt werden. Kapitel 4 gibt durch eine Eigenwertanalyse weitere Einblicke in das Innenleben der verzögerungsbasierten Reservoirberechnung. Diese Analyse bekräftigt die Ergebnisse aus Kapitel 3 durch einen mathematischen Linearisierungsansatz. Die Imaginärteile der Eigenwerte weisen darauf hin, dass Resonanzen vermieden werden sollten, da sie den nutzbaren Phasenraum für den Reservoir-Computer verringern. Währenddessen zeigen die Realteile, dass Betriebspunkte nahe der Kritikalität für Aufgaben, die einen Langzeitspeicher erfordern, am günstigsten sind, während Betriebspunkte, die von der Kritikalität entfernt sind, schnellere, stärker nichtlineare Systeme mit weniger Speicher ergeben. Kapitel 5 stellt eine analytische Formel für die lineare Speicherkapazität eines beliebigen eindimensionalen verzögerungsbasierten Reservoir-Computers vor, die als „Master Memory Function" bezeichnet wird. Dies ist ein wichtiges Ergebnis dieser Arbeit, da es die Vorhersage der linearen Speicherkapazität von verzögerungsbasierten Reservoir-Computern für Systeme mit unbekannten Modellen anhand von Messungen ermöglicht. Die „Master Memory Function" weist auch eine universelle Eigenschaft auf: Alle verzögerungsbasierten Reservoir-Computer, die von einem kleinen Eingangssignal angetrieben werden, sodass ihre jeweiligen Linearisierungen dasselbe System ergeben, auch dieselbe lineare Speicherkapazität haben. Die Möglichkeit, die Formel auf mehrere zeitverzögerte Rückkopplungen und höhere Dimensionalität zu erweitern, wird erörtert, und interessierten Lesern wird eine weitere Erforschung empfohlen. Kapitel 6 vergleicht verzögerungsbasierte Reservoir-Computer mit anderen weit verbreiteten Ansätzen zur Zeitserienprognose, insbesondere dem nichtlinearen autoregresiven Modell und dem Kernel-Trick im reproduzierenden Kernel-Hilbert-Raum. Es wird gezeigt, dass Reservoir-Computer in der Lage sind, bei Aufgaben mit besonders langen Korrelationen Vorhersagen von vergleichbarer Qualität zu machen, und zwar aufgrund des inhärenten Speichers, der reservoir-basierten Ansätzen eigen ist. Im abschließenden Kapitel 7 wird ein hybrider Ansatz vorgestellt, mit dem jeder Aspekt des Reservoir-Computer-Schemas verbessert wird. Eine datengesteuerte Prognosemethode namens SINDy wird auf denselben Datensatz angewendet, um dynamische Modelle abzuleiten. Diese Modelle werden in Verbindung mit einem verzögerungsbasierten Reservoir-Computer verwendet, was zu einer erheblichen Verbesserung der kurzfristigen Vorhersagegenauigkeit und der langfristigen Vorhersagestabilität führt, während gleichzeitig die Komplexität der benötigten Modelle reduziert wird. Zusammenfassend lässt sich sagen, dass diese Arbeit einen Einblick in das Innenleben der verzögerungsbasierten Reservoir-Computer gibt, bei der Ermittlung optimaler Betriebspunkte hilft und eine analytische Beschreibung der linearen Speicherkapazität liefert. Der verzögerungsbasierte Reservoir-Computer wird in den breiteren Rahmen der datengesteuerten Zeitserienprognose eingeordnet und führt das Konzept eines hybriden datengesteuerten Ansatzes in die Reservoir-Computer-Gemeinschaft ein, wobei vielversprechende Ergebnisse erzielt werden.
- Published
- 2023
- Full Text
- View/download PDF
18. Data-informed reservoir computing for efficient time-series prediction.
- Author
-
Köster, Felix, Patel, Dhruvit, Wikner, Alexander, Jaurigue, Lina, and Lüdge, Kathy
- Subjects
- *
NONLINEAR dynamical systems , *MACHINE learning , *DYNAMICAL systems - Abstract
We propose a new approach to dynamical system forecasting called data-informed-reservoir computing (DI-RC) that, while solely being based on data, yields increased accuracy, reduced computational cost, and mitigates tedious hyper-parameter optimization of the reservoir computer (RC). Our DI-RC approach is based on the recently proposed hybrid setup where a knowledge-based model is combined with a machine learning prediction system, but it replaces the knowledge-based component by a data-driven model discovery technique. As a result, our approach can be chosen when a suitable knowledge-based model is not available. We demonstrate our approach using a delay-based RC as the machine learning component in conjunction with sparse identification of nonlinear dynamical systems for the data-driven model component. We test the performance on two example systems: the Lorenz system and the Kuramoto–Sivashinsky system. Our results indicate that our proposed technique can yield an improvement in the time-series forecasting capabilities compared with both approaches applied individually, while remaining computationally cheap. The benefit of our proposed approach, compared with pure RC, is most pronounced when the reservoir parameters are not optimized, thereby reducing the need for hyperparameter optimization. [ABSTRACT FROM AUTHOR]
- Published
- 2023
- Full Text
- View/download PDF
19. Deriving task specific performance from the information processing capacity of a reservoir computer.
- Author
-
Hülser, Tobias, Köster, Felix, Lüdge, Kathy, and Jaurigue, Lina
- Subjects
PROCESS capability ,TASK performance ,INFORMATION processing ,NONLINEAR oscillators ,TEST methods - Abstract
In the reservoir computing literature, the information processing capacity is frequently used to characterize the computing capabilities of a reservoir. However, it remains unclear how the information processing capacity connects to the performance on specific tasks. We demonstrate on a set of standard benchmark tasks that the total information processing capacity correlates poorly with task specific performance. Further, we derive an expression for the normalized mean square error of a task as a weighted function of the individual information processing capacities. Mathematically, the derivation requires the task to have the same input distribution as used to calculate the information processing capacities. We test our method on a range of tasks that violate this requirement and find good qualitative agreement between the predicted and the actual errors as long as the task input sequences do not have long autocorrelation times. Our method offers deeper insight into the principles governing reservoir computing performance. It also increases the utility of the evaluation of information processing capacities, which are typically defined on i.i.d. input, even if specific tasks deliver inputs stemming from different distributions. Moreover, it offers the possibility of reducing the experimental cost of optimizing physical reservoirs, such as those implemented in photonic systems. [ABSTRACT FROM AUTHOR]
- Published
- 2023
- Full Text
- View/download PDF
20. Role of delay-times in delay-based photonic reservoir computing [Invited]
- Author
-
Hülser, Tobias, primary, Köster, Felix, additional, Jaurigue, Lina, additional, and Lüdge, Kathy, additional
- Published
- 2022
- Full Text
- View/download PDF
21. Collective Coherence Resonance in Networks of Optical Neurons
- Author
-
Köster, Felix, primary, Lingnau, Benjamin, additional, Krimlowski, Andrej, additional, Hövel, Philipp, additional, and Lüdge, Kathy, additional
- Published
- 2021
- Full Text
- View/download PDF
22. Deep Time-Delay Reservoir Computing: Dynamics and Memory Capacity
- Author
-
Goldmann, Mirko, Köster, Felix, Lüdge, Kathy, Yanchuk, Serhiy, Goldmann, Mirko, Köster, Felix, Lüdge, Kathy, and Yanchuk, Serhiy
- Published
- 2021
23. Früherkennung von psychischer Komorbidität in der stationären dermatologischen und internistischen Versorgung: Darstellung eines neuen Versorgungskonzeptes
- Author
-
Kohlmann, Sebastian, additional, Köster, Felix-Wilhelm, additional, Braunschneider, Lea-Elena, additional, Meier, Anja Hermann, additional, Lohse, Ansgar W., additional, Schneider, Stefan W., additional, Loeper, Siobhan, additional, and Löwe, Bernd, additional
- Published
- 2021
- Full Text
- View/download PDF
24. Insight into delay based reservoir computing via eigenvalue analysis
- Author
-
Köster, Felix, primary, Yanchuk, Serhiy, additional, and Lüdge, Kathy, additional
- Published
- 2021
- Full Text
- View/download PDF
25. Screening auf psychische Komorbiditäten in der Dermatologie
- Author
-
Köster, Felix-Wilhelm, primary, Kohlmann, Sebastian, additional, Loeper, Siobhan, additional, Löwe, Bernd, additional, and Schneider, Stefan W., additional
- Published
- 2020
- Full Text
- View/download PDF
26. Limitations of the Recall Capabilities in Delay-Based Reservoir Computing Systems
- Author
-
Köster, Felix, primary, Ehlert, Dominik, additional, and Lüdge, Kathy, additional
- Published
- 2020
- Full Text
- View/download PDF
27. Deep time-delay reservoir computing: Dynamics and memory capacity
- Author
-
Goldmann, Mirko, primary, Köster, Felix, additional, Lüdge, Kathy, additional, and Yanchuk, Serhiy, additional
- Published
- 2020
- Full Text
- View/download PDF
28. Temperature dependent linewidth rebroadening in quantum dot semiconductor lasers
- Author
-
Köster, Felix, primary, Duan, Jianan, additional, Dong, Bozhang, additional, Huang, Heming, additional, Grillot, Frédéric, additional, and Lüdge, Kathy, additional
- Published
- 2020
- Full Text
- View/download PDF
29. Berücksichtigung von Verkehrsmanagement Information in der Logistik Planung
- Author
-
Köster, Felix and Mattfeld, Dirk C.
- Subjects
doctoral thesis ,ddc:0 ,ddc:00 ,ddc:33 ,ddc:006 ,ddc:3 - Abstract
Today, logistics and traffic management are two important elements for the urban population. Logistic service providers enable the availability of consumer goods to city residents and the traffic management optimizes traffic flows of the city by operating the traffic infrastructure (i.e. traffic signals). This dissertation investigates if logistics and traffic management can achieve a mutual benefit through an information sharing cooperation. The main motivation for a logistics company, which operates in the high cost pressure delivery industry, is a possible reduction of operational costs for their delivery fleet. The motivation of the traffic management to enter a cooperation is a possible reduction of the high traffic burden induced by freight vehicles. In such a cooperation the traffic management could transfer information about the state of the operable traffic infrastructure, and therefore their related current travel time impact, to the logistics company. With this information the logistics company can optimize their delivery tours. This dissertation assesses the benefits of cooperation by analyzing how the delivery tours can be improved and how the freight traffic burden can be reduced. This question is investigated in a case study of an environmental-sensitive traffic management system in the city of Braunschweig, Germany and a city logistics service provider. The air pollution in Braunschweig does not comply with the EU air pollution limit. If the air pollution is high in a hot-spot area, the traffic management changes the traffic infrastructure to reduce the traffic flow into the hot-spot areas, which affects the surrounding traffic situation. In the case study, the degree of information from the traffic management and the tasks of the logistics company are varied and formulated in different dynamic Vehicle Routing Problems. The evaluation of the case study shows that a mutual benefit of the cooperation can be discovered over all variations of the experiments. Der kontinuierlich wachsende E-Commerce führt weltweit zu in einem erhöhten innerstädtischen Frachtverkehr. Dieser Frachtverkehr wird von Logistikdienstleister durchgeführt, die einzelne Frachtsendungen zu Auslieferungstouren bündeln. Aufgrund von schwankenden Fahrzeiten im innerstädtischen Umfeld ist eine effiziente Planung der Auslieferungstouren für die Lieferfahrzeuge jeodch schwierig. Um die Verkehrssituation zu verbessern setzten viele Städte Verkehrsmanagementsysteme ein. Diese optimieren den Verkehr durch die Anpassung der Verkehrsleitstrategie an die Verkehrssituation. Diese Dissertation untersucht, ob eine Kooperation zum Austausch von Informationen zwischen einem Logistikdienstleister und dem Verkehrsmanagement die jeweiligen Interessen verbessert. Aufgrund des hohen Kostendrucks der Logistikbranche ist ein Logistikdienstleister an der Verkehrsleitstrategie interessiert um damit die Kosten ihrer Touren zu senken. Das Verkehrsmanagement erhofft sich eine Verminderung der Verkehrsbelastung durch den Frachtverkehr. Der Nutzen einer Kooperation wird in einer Fallstudie mit dem umweltorientierten Verkehrsmanagement in Braunschweig und einem Paketdienstleister analysiert. Dafür wird die Aufgabenstellung des Paketdienstleisters als dynamisches Vehicle Routing Problem, in dem das Verkehrsmanagement die Fahrzeiten beeinflusst, formuliert und mit unterschiedlichen Methoden gelöst. In den Experimenten wird der Einfluss der Verkehrsleitstrategien auf die Fahrzeiten, die Anzahl der Fahrzeuge in der Logistikflotte und zwei Variationen des dynamischen Vehicle Routing Problems betrachtet. Aus den Ergebnissen ist ersichtlich, dass eine Kooperation zwischen Logistikdienstleister und Verkehrsmanagement als sinnvoll zu erachten ist, da sie zu einer Reduktion in der Verkehrsbelastung durch Frachtverkehr in kritischen Bereichen und zu einer Reduktion der Fahrzeiten der Auslieferungstouren führt.
- Published
- 2017
30. [Early Detection of Psychological Comorbidity in Patients Admitted to Dermatological and Internal Medicine Wards: A New Care Model for Psychosomatic Consultation Service].
- Author
-
Kohlmann S, Köster FW, Braunschneider LE, Meier AH, Lohse AW, Schneider SW, Loeper S, and Löwe B
- Subjects
- Anxiety Disorders, Comorbidity, Early Diagnosis, Humans, Psychophysiologic Disorders diagnosis, Psychophysiologic Disorders epidemiology, Psychophysiologic Disorders therapy, Referral and Consultation
- Abstract
This article explains the development and implementation of a psychosomatic screening and consultation service for inpatient somatic care. Approximately one in six somatic inpatients has a mental disorder. It is estimated that only half of these cases are properly identified. Consequently, a large proportion of patients remains untreated. To address this gap in care, a psychosomatic early detection programme was developed by an interdisciplinary working group. This programme is based on the Patient Health Questionnaire-4 (PHQ-4), a psychometrically very well evaluated ultra-short screening questionnaire for the detection of depressive and anxiety disorders. For implementation in routine inpatient care, the PHQ-4 was programmed as a form in the electronic medical record and administered by nursing staff during the admission interview. If the PHQ-4 screening result indicates the presence of a mental comorbidity and the patient expresses a wish for assessment of this disorder, a psychosomatic consultation is automatically ordered. The PHQ-4 was implemented into the clinical routine in four internal medicine and three dermatology wards of the University Medical Center Hamburg-Eppendorf. Documentation of the early diagnosis in the electronic patient record is a minimally costly, less time-consuming and practicable method of providing patients with holistic care through rapid interdisciplinary referral. An evaluation of cost-effectiveness, clinical efficiency, and acceptance is still pending., Competing Interests: Die Autorinnen/Autoren geben an, dass kein Interessenkonflikt besteht., (Thieme. All rights reserved.)
- Published
- 2021
- Full Text
- View/download PDF
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.