TY - CONF AU - Abbas, Nilab AU - Bauer, Anna Brigitte AU - Reinhold, Peter ED - van Vorst, Helena ID - 45759 T2 - Lernen, Lehren und Forschen in einer digital geprägten Welt TI - PSΦ: Entwicklung von Unterstützungsmaßnahmen für Theoretische Physik ER - TY - CONF AU - Bauer, Anna Brigitte AU - Reinhold, Peter ED - van Vorst, Helena ID - 45758 T2 - Lernen, Lehren und Forschen in einer digital geprägten Welt TI - PSФ: Entwicklung einer abgestimmten Studieneingangsphase (Physik) ER - TY - JOUR AB - A series of new organic donor–π–acceptor dyes incorporating a diquat moiety as a novel electron-acceptor unit have been synthesized and characterized. The analytical data were supported by DFT calculations. These dyes were explored in the aerobic thiocyanation of indoles and pyrroles. Here they showed a high photocatalytic activity under visible light, giving isolated yields of up to 97 %. In addition, the photocatalytic activity of standalone diquat and methyl viologen through formation of an electron donor acceptor complex is presented. AU - Meier, Armin AU - Badalov, Sabuhi AU - Biktagirov, Timur AU - Schmidt, Wolf Gero AU - Wilhelm, René ID - 43827 IS - 22 JF - Chemistry – A European Journal KW - General Chemistry KW - Catalysis KW - Organic Chemistry SN - 0947-6539 TI - Diquat Based Dyes: A New Class of Photoredox Catalysts and Their Use in Aerobic Thiocyanation VL - 29 ER - TY - JOUR AB - As a benchmark, the structural, electronic and optical properties of the three main phases of TiO$\rm{_2}$ crystals have been calculated using Hubbard U correction and hybrid functional methods in density-functional theory. These calculations are compared concerning the available experimental observations on pristine TiO$\rm{_2}$ crystals. Modified hybrid functionals, particularly the PBE0 functional with 11.4% fraction of exact exchange, are shown to provide highly accurate atomic structures and also accurate electronic structure data, including optical excitation energies. With $\rm{DFT+U}$, accurate optical spectra are also possible, but only if the Hubbard U is applied on the O $\rm2p$ electrons exclusively. Furthermore, both methods, the 11.4%-PBE0 hybrid functional and the $\rm{DFT+U_p}$ scheme have been used to study TiO$\rm{_2}$ amorphous ultra-thin films, confirming the agreement of the two methods even with respect to small details of the optical spectra. Our results show that the proposed $\rm{DFT+U_p}$ methodology is computationally efficient, but still accurate. It can be applied to well-ordered TiO$\rm{_2}$ polymorphs as well as to amorphous TiO$\rm{_2}$ and will allow for the calculations of complex titania-based structures. AU - Badalov, Sabuhi AU - Bocchini, Adriana AU - Wilhelm, Rene AU - Kozub, A. L. AU - Gerstmann, Uwe AU - Schmidt, Wolf Gero ID - 45764 JF - Materials Research Express TI - Rutile, anatase, brookite and titania thin film from Hubbard corrected and hybrid DFT ER - TY - JOUR AB - The development of automotive components with reduced greenhouse gas (GHG) emissions is needed to reduce overall vehicle emissions. Life Cycle Engineering (LCE) based on Life Cycle Assessment (LCA) supports this by providing holistic information and improvement potentials regarding eco-efficient products. Key factors influencing LCAs of automotive components, such as material production, will change in the future. First approaches for integrating future scenarios for these key factors into LCE already exist, but they only consider a limited number of parameters and scenarios. This work aims to develop a method that can be practically applied in the industry for integrating prospective LCAs (pLCA) into the LCE of automotive components, considering relevant parameters and consistent scenarios. Therefore, pLCA methods are further developed to investigate the influence of future scenarios on the GHG emissions of automotive components. The practical application is demonstrated for a vehicle component with different design options. This paper shows that different development paths of the foreground and background system can shift the ecological optimum of design alternatives. Therefore, future pathways of relevant parameters must be considered comprehensively to reduce GHG emissions of future vehicles. This work contributes to the methodological and practical integration of pLCA into automotive development processes and provides quantitative results. AU - Grenz, Julian AU - Ostermann, Moritz AU - Käsewieter, Karoline AU - Cerdas, Felipe AU - Marten, Thorsten AU - Herrmann, Christoph AU - Tröster, Thomas ID - 45782 IS - 13 JF - Sustainability KW - prospective LCA KW - life cycle engineering (LCE) KW - lightweight design KW - automotive components KW - body parts KW - circular economy KW - steel KW - aluminum KW - hybrid materials KW - fiber metal laminates SN - 2071-1050 TI - Integrating Prospective LCA in the Development of Automotive Components VL - 15 ER - TY - CONF AB - The notion of neural collapse refers to several emergent phenomena that have been empirically observed across various canonical classification problems. During the terminal phase of training a deep neural network, the feature embedding of all examples of the same class tend to collapse to a single representation, and the features of different classes tend to separate as much as possible. Neural collapse is often studied through a simplified model, called the unconstrained feature representation, in which the model is assumed to have "infinite expressivity" and can map each data point to any arbitrary representation. In this work, we propose a more realistic variant of the unconstrained feature representation that takes the limited expressivity of the network into account. Empirical evidence suggests that the memorization of noisy data points leads to a degradation (dilation) of the neural collapse. Using a model of the memorization-dilation (M-D) phenomenon, we show one mechanism by which different losses lead to different performances of the trained network on noisy data. Our proofs reveal why label smoothing, a modification of cross-entropy empirically observed to produce a regularization effect, leads to improved generalization in classification tasks. AU - Nguyen, Duc Anh AU - Levie, Ron AU - Lienen, Julian AU - Kutyniok, Gitta AU - Hüllermeier, Eyke ID - 31880 T2 - International Conference on Learning Representations, ICLR TI - Memorization-Dilation: Modeling Neural Collapse Under Noise ER - TY - GEN AB - Label noise poses an important challenge in machine learning, especially in deep learning, in which large models with high expressive power dominate the field. Models of that kind are prone to memorizing incorrect labels, thereby harming generalization performance. Many methods have been proposed to address this problem, including robust loss functions and more complex label correction approaches. Robust loss functions are appealing due to their simplicity, but typically lack flexibility, while label correction usually adds substantial complexity to the training setup. In this paper, we suggest to address the shortcomings of both methodologies by "ambiguating" the target information, adding additional, complementary candidate labels in case the learner is not sufficiently convinced of the observed training label. More precisely, we leverage the framework of so-called superset learning to construct set-valued targets based on a confidence threshold, which deliver imprecise yet more reliable beliefs about the ground-truth, effectively helping the learner to suppress the memorization effect. In an extensive empirical evaluation, our method demonstrates favorable learning behavior on synthetic and real-world noise, confirming the effectiveness in detecting and correcting erroneous training labels. AU - Lienen, Julian AU - Hüllermeier, Eyke ID - 45814 T2 - arXiv:2305.13764 TI - Mitigating Label Noise through Data Ambiguation ER - TY - CONF AB - Many applications require explainable node classification in knowledge graphs. Towards this end, a popular ``white-box'' approach is class expression learning: Given sets of positive and negative nodes, class expressions in description logics are learned that separate positive from negative nodes. Most existing approaches are search-based approaches generating many candidate class expressions and selecting the best one. However, they often take a long time to find suitable class expressions. In this paper, we cast class expression learning as a translation problem and propose a new family of class expression learning approaches which we dub neural class expression synthesizers. Training examples are ``translated'' into class expressions in a fashion akin to machine translation. Consequently, our synthesizers are not subject to the runtime limitations of search-based approaches. We study three instances of this novel family of approaches based on LSTMs, GRUs, and set transformers, respectively. An evaluation of our approach on four benchmark datasets suggests that it can effectively synthesize high-quality class expressions with respect to the input examples in approximately one second on average. Moreover, a comparison to state-of-the-art approaches suggests that we achieve better F-measures on large datasets. For reproducibility purposes, we provide our implementation as well as pretrained models in our public GitHub repository at https://github.com/dice-group/NeuralClassExpressionSynthesis AU - KOUAGOU, N'Dah Jean AU - Heindorf, Stefan AU - Demir, Caglar AU - Ngonga Ngomo, Axel-Cyrille ED - Pesquita, Catia ED - Jimenez-Ruiz, Ernesto ED - McCusker, Jamie ED - Faria, Daniel ED - Dragoni, Mauro ED - Dimou, Anastasia ED - Troncy, Raphael ED - Hertling, Sven ID - 33734 KW - Neural network KW - Concept learning KW - Description logics T2 - The Semantic Web - 20th Extended Semantic Web Conference (ESWC 2023) TI - Neural Class Expression Synthesis VL - 13870 ER - TY - GEN AB - Label noise poses an important challenge in machine learning, especially in deep learning, in which large models with high expressive power dominate the field. Models of that kind are prone to memorizing incorrect labels, thereby harming generalization performance. Many methods have been proposed to address this problem, including robust loss functions and more complex label correction approaches. Robust loss functions are appealing due to their simplicity, but typically lack flexibility, while label correction usually adds substantial complexity to the training setup. In this paper, we suggest to address the shortcomings of both methodologies by "ambiguating" the target information, adding additional, complementary candidate labels in case the learner is not sufficiently convinced of the observed training label. More precisely, we leverage the framework of so-called superset learning to construct set-valued targets based on a confidence threshold, which deliver imprecise yet more reliable beliefs about the ground-truth, effectively helping the learner to suppress the memorization effect. In an extensive empirical evaluation, our method demonstrates favorable learning behavior on synthetic and real-world noise, confirming the effectiveness in detecting and correcting erroneous training labels. AU - Lienen, Julian AU - Hüllermeier, Eyke ID - 45244 T2 - arXiv:2305.13764 TI - Mitigating Label Noise through Data Ambiguation ER - TY - JOUR AB - Perfect vector vortex beams (PVVBs) have attracted considerable interest due to their peculiar optical features. PVVBs are typically generated through the superposition of perfect vortex beams, which suffer from the limited number of topological charges (TCs). Furthermore, dynamic control of PVVBs is desirable and has not been reported. We propose and experimentally demonstrate hybrid grafted perfect vector vortex beams (GPVVBs) and their dynamic control. Hybrid GPVVBs are generated through the superposition of grafted perfect vortex beams with a multifunctional metasurface. The generated hybrid GPVVBs possess spatially variant rates of polarization change due to the involvement of more TCs. Each hybrid GPVVB includes different GPVVBs in the same beam, adding more design flexibility. Moreover, these beams are dynamically controlled with a rotating half waveplate. The generated dynamic GPVVBs may find applications in the fields where dynamic control is in high demand, including optical encryption, dense data communication, and multiple particle manipulation. AU - Ahmed, Hammad AU - Ansari, Muhammad Afnan AU - Li, Yan AU - Zentgraf, Thomas AU - Mehmood, Muhammad Qasim AU - Chen, Xianzhong ID - 45868 IS - 1 JF - Nature Communications KW - General Physics and Astronomy KW - General Biochemistry KW - Genetics and Molecular Biology KW - General Chemistry KW - Multidisciplinary SN - 2041-1723 TI - Dynamic control of hybrid grafted perfect vector vortex beams VL - 14 ER - TY - CHAP AU - Götte, Thorsten AU - Knollmann, Till AU - Meyer auf der Heide, Friedhelm AU - Scheideler, Christian AU - Werthmann, Julian ED - Haake, Claus-Jochen ED - Meyer auf der Heide, Friedhelm ED - Platzner, Marco ED - Wachsmuth, Henning ED - Wehrheim, Heike ID - 45875 T2 - On-The-Fly Computing -- Individualized IT-services in dynamic markets TI - Capabilities and Limitations of Local Strategies in Dynamic Networks VL - 412 ER - TY - CHAP AU - Karl, Holger AU - Maack, Marten AU - Meyer auf der Heide, Friedhelm AU - Pukrop, Simon AU - Redder, Adrian ED - Haake, Claus-Jochen ED - Meyer auf der Heide, Friedhelm ED - Platzner, Marco ED - Wachsmuth, Henning ED - Wehrheim, Heike ID - 45895 T2 - On-The-Fly Computing -- Individualized IT-services in dynamic markets TI - On-The-Fly Compute Centers II: Execution of Composed Services in Configurable Compute Centers VL - 412 ER - TY - CHAP AU - Blömer, Johannes AU - Bobolz, Jan AU - Eidens, Fabian AU - Jager, Tibor AU - Kramer, Paul ED - Haake, Claus-Jochen ED - Meyer auf der Heide, Friedhelm ED - Platzner, Marco ED - Wachsmuth, Henning ED - Wehrheim, Heike ID - 45901 T2 - On-The-Fly Computing -- Individualized IT-services in dynamic markets TI - Practical Cryptograhic Techniques for Secure and Privacy-Preserving Customer Loyalty Systems VL - 412 ER - TY - CHAP AU - Boschmann, Alexander AU - Clausing, Lennart AU - Jentzsch, Felix AU - Ghasemzadeh Mohammadi, Hassan AU - Platzner, Marco ED - Haake, Claus-Jochen ED - Meyer auf der Heide, Friedhelm ED - Platzner, Marco ED - Wachsmuth, Henning ED - Wehrheim, Heike ID - 45899 T2 - On-The-Fly Computing -- Individualized IT-services in dynamic markets TI - Flexible Industrial Analytics on Reconfigurable Systems-On-Chip VL - 412 ER - TY - CHAP AU - Wehrheim, Heike AU - Platzner, Marco AU - Bodden, Eric AU - Schubert, Philipp AU - Pauck, Felix AU - Jakobs, Marie-Christine ED - Haake, Claus-Jochen ED - Meyer auf der Heide, Friedhelm ED - Platzner, Marco ED - Wachsmuth, Henning ED - Wehrheim, Heike ID - 45888 T2 - On-The-Fly Computing -- Individualized IT-services in dynamic markets TI - Verifying Software and Reconfigurable Hardware Services VL - 412 ER - TY - CHAP AU - Gottschalk, Sebastian AU - Vorbohle, Christian AU - Kundisch, Dennis AU - Engels, Gregor AU - Wünderlich, Nacy V. ED - Haake, Claus-Jochen ED - Meyer auf der Heide, Friedhelm ED - Platzner, Marco ED - Wachsmuth, Henning ED - Wehrheim, Heike ID - 45897 T2 - On-The-Fly Computing -- Individualized IT-services in dynamic markets TI - Architectural Management of OTF Computing Markets VL - 412 ER - TY - CHAP AU - Blömer, Johannes AU - Eidens, Fabian AU - Jager, Tibor AU - Niehues, David AU - Scheideler, Christian ED - Haake, Claus-Jochen ED - Meyer auf der Heide, Friedhelm ED - Platzner, Marco ED - Wachsmuth, Henning ED - Wehrheim, Heike ID - 45891 T2 - On-The-Fly Computing -- Individualized IT-services in dynamic markets TI - Robustness and Security VL - 412 ER - TY - CHAP AU - Bäumer, Frederik Simon AU - Chen, Wei-Fan AU - Geierhos, Michaela AU - Kersting, Joschka AU - Wachsmuth, Henning ED - Haake, Claus-Jochen ED - Meyer auf der Heide, Friedhelm ED - Platzner, Marco ED - Wachsmuth, Henning ED - Wehrheim, Heike ID - 45882 T2 - On-The-Fly Computing -- Individualized IT-services in dynamic markets TI - Dialogue-based Requirement Compensation and Style-adjusted Data-to-text Generation VL - 412 ER - TY - CHAP AU - Hanselle, Jonas Manuel AU - Hüllermeier, Eyke AU - Mohr, Felix AU - Ngonga Ngomo, Axel-Cyrille AU - Sherif, Mohamed AU - Tornede, Alexander AU - Wever, Marcel Dominik ED - Haake, Claus-Jochen ED - Meyer auf der Heide, Friedhelm ED - Platzner, Marco ED - Wachsmuth, Henning ED - Wehrheim, Heike ID - 45884 T2 - On-The-Fly Computing -- Individualized IT-services in dynamic markets TI - Configuration and Evaluation VL - 412 ER - TY - CHAP AU - Haake, Claus-Jochen AU - Hehenkamp, Burkhard AU - Polevoy, Gleb ED - Haake, Claus-Jochen ED - Meyer auf der Heide, Friedhelm ED - Platzner, Marco ED - Wachsmuth, Henning ED - Wehrheim, Heike ID - 45878 T2 - On-The-Fly Computing -- Individualized IT-services in dynamic markets TI - The Market for Services: Incentives, Algorithms, Implementation VL - 412 ER -