107036 research outputs found
Sort by
Unraveling European electricity price volatility: the impact of renewables
No full textThe transition from fossil fuels to renewable, carbon-free energy is a critical challenge for mitigating climate change, with different European countries adopting diverse approaches to integrating renewable energy sources into their energy systems. Hereby, the intermittent nature and variability of renewable energy generation lead to significant challenges to electricity markets, amplifying supply volatility and market instability. This study examines the determinants of electricity price volatility in European wholesale markets, focusing on the impact of renewable energy sources as well as the influence of fuel prices and electricity demand. Using daily data from thirteen European electricity markets for the period 2015 to 2023, we account for the dynamic development of the energy system and volatile market condition, including the recent shocks induced by the Russia–Ukraine war. By employing country-specific and panel regression models, the findings reveal that conventional power sources, such as nuclear and coal, tend to stabilize electricity markets. In contrast, a higher share of solar power is associated with increased risk. Wind power exhibits mixed effects, with stabilizing impacts in most markets and volatility-enhancing tendencies in Spain, highlighting the role of market-specific dynamics. Additionally, total electricity load as well as the Russia-Ukraine war emerge as significant drivers of market risk
Responsible AI im Case Management: zur Gestaltung von verantwortungsvollen Praktiken in soziotechnischen Handlungsdynamiken
No full textLearning approaches for interactive segmentation: from unimodal to multimodal interactive segmentation
No full textOne of the most important tasks computer vision has automated is the exact localization of objects in images. When we consider the notion of an exact location of an object in an image, we would like to know precisely which pixels are occupied by the object and which are not. This pixel-precise localization is referred to as image segmentation in the context of computer vision. In recent years, deep neural networks have become the central part of most systems for image segmentation. Although the methods and architectures for the various image segmentation tasks have improved over the last years, training neural networks for such systems still relies on the availability of large amounts of annotated data. Specifically for segmentation tasks, annotated data is hard to acquire. While the images themselves are usually easy to procure at a large scale, manually creating the masks for surfaces in the image requires a considerable amount of effort. To ease the process of manually creating segmentation masks, interactive segmentation systems have been created. These systems allow the user to place clicks on the image and then try to automatically infer a high quality mask on the basis of these clicks.
In this thesis, we investigate the topic of interactive segmentation systems that are based on neural networks. In the first part of this thesis, we focus on architectures which predict segmentation masks on the basis of RGB images. Therein, we first propose a method which allows the network to continue learning while it is currently in use. To do so, we only use information that is generated as a byproduct of using the model. Afterwards, we propose a novel network architecture for interactive segmentation. We design the architecture in such a way that it allows for quick responses after each click.
In the second part of this thesis, we leverage other modalities than RGB images to improve the segmentation performance of our networks. We use the geometric information from depth maps as an additional input modality alongside the RGB images, which results in better segmentation masks. Since corresponding depth maps are not generally available for arbitrary images, we generate pseudo depth maps using networks that have been pretrained for the task of monocular depth estimation. Even when replacing RGB images with high-quality depth maps as an input modality, we observe performance increases for some scenarios. We also develop a novel architecture that is capable of integrating information from an arbitrary amount of modalities. The multi-modal fusion strategy is designed to allow for the usage of an inaccessible gray-box feature extractor for RGB images. On top of this, we propose an extended version of the evaluation mechanism for interactive segmentation that accounts for challenges that occur when we want to segment multiple surfaces in the same image
Classification and stability of penalized pinned elasticae
No full textThis paper considers critical points of the length-penalized elastic bending energy among planar curves whose endpoints are fixed. We classify all critical points with an explicit parametrization. The classification strongly depends on a special penalization parameter λˆ ≃ 0.70107. Stability of all the critical points is also investigated, and again the threshold λˆ plays a decisive role. In addition, our explicit parametrization is applied to compare the energy of critical points, leading to uniqueness of minimal nontrivial critical points.
As an application we obtain eventual embeddedness of elastic flows
Determinants and forecasting of corporate greenwashing behavior
No full textThis paper empirically analyzes the determinants of corporate greenwashing behavior to enhance forecasting and mitigation of greenwashing practices, particularly in the context of stakeholder decision-making. Using company-level characteristics from a sample of STOXX Europe 600 constituents, we show that ESG and environmental (E) scores exhibit a U-shaped relationship with greenwashing, indicating that companies with both low and high (E)SG scores are more likely to engage in greenwashing. Additionally, ESG disclosure score, company size, cash-to-assets, and capital intensity are positively associated with greenwashing behavior. Furthermore, greenwashing behavior is more prevalent in consumer-related industries than in other industries. Building on the identified determinants of greenwashing behavior, we develop machine learning models grounded in economic theory to forecast greenwashing risk. Overall, our analyses demonstrate how current and future greenwashing risks can be effectively assessed. This enables stakeholders such as investors and policymakers to better identify corporate greenwashing behavior and incorporate the associated risks into their decision-making
Krankheit als Weg zur Dichtung? Bertolt Brecht, die Mutter, seine Anamnese und frühe Krankheitsgeschichte
No full textComparative performance of the nutritional risk index and modified 5-item frailty index in predicting surgical outcomes: a multi-institutional analysis of 9.7 million patients
No full text
Die digitale Transformation des Tourismus: akteurszentrierte Perspektiven auf Smart Tourism in urbanen Räumen
No full textDie Digitalisierung zählt zu den bedeutendsten gesellschaftlichen Transformationsprozessen der Gegenwart und beeinflusst nahezu alle Lebensbereiche. Auch im Tourismus findet eine entsprechende Transformation statt. So bezeichnet Smart Tourism den Einsatz von Informations- und Kommunikationstechnologien in touristischen Prozessen zur Optimierung von Effizienz, Wettbewerbsfähigkeit und Erlebnisqualität. Die bisherige Forschung bzgl. dieses Transformationsprozesses fokussiert sich jedoch überwiegend auf technologische Aspekte und theoretische Konzeptualisierungen, während sozialwissenschaftliche Perspektiven – insbesondere die Rolle der beteiligten Akteure – bislang kaum systematisch analysiert wurden. Vor diesem Hintergrund untersucht die vorliegende Arbeit, inwiefern verschiedene Akteursgruppen die Entwicklung und Umsetzung von Smart Tourism in urbanen Destinationen beeinflussen. Auf Basis der Strukturationstheorie nach Giddens werden dazu in einem Mixed-Methods-Ansatz Governance-Akteure, touristische Leistungsträger und TouristInnen sowie deren Handlungs- und Interaktionsmuster in bayerischen Städten empirisch analysiert und Typen von Anbietern und Nachfragern identifiziert. Die Untersuchungsergebnisse zeigen, dass Smart Tourism nicht maßgeblich durch technologische Innovationen bestimmt wird, sondern vor allem durch die Handlungen und Entscheidungen der beteiligten Akteure, deren Vielfalt und unterschiedliche Strategien sowie Interaktionen den Erfolg von Smart Tourism im urbanen Kontext entscheidend beeinflussen. Diese Erkenntnisse werden in einem neu entwickelten Actor-Driven Multidimensional Transformation Model (AMTM) zusammengeführt, das es ermöglicht zu analysieren, wie die Akteure den Transformationsprozess gestalten und strukturelle Veränderungen anstoßen oder bestehende Strukturen stabilisieren. Darüber hinaus dient das AMTM als Grundlage für die Entwicklung eines akteurszentrierten Indikatorensystems zur differenzierten Erfassung von Smart Tourism in der Praxis. Die Arbeit leistet damit einen wesentlichen Beitrag zur Erforschung digitalisierungsbedingter Transformationsprozesse im Tourismus, indem sie den wissenschaftlichen Diskurs um eine sozialwissenschaftliche Perspektive erweitert und aufzeigt, dass es sich bei Smart Tourism um einen soziotechnischen Veränderungsprozess handelt, bei dem das Zusammenspiel von Technologie und Akteurshandeln zentral ist