Hochschule Konstanz University of Applied Sciences
Hochschulschriftenserver der HTWG KonstanzNot a member yet
2319 research outputs found
Sort by
Hochwertiges Recycling von Post-Consumer-Flachglasabfällen aus dem Gebäudesektor in Baden-Württemberg
No full textDer Bericht beschreibt den Lebenszyklus des Flachglases. Des Weiteren wird aufgezeigt, welche Maßnahmen nötig sind, um die Menge an Flachglasscherben einem hochwertigen Recycling zuzuführen. Kern des Berichtes sind die Auswertung von Experteninterviews. Daran nahmen Expertinnen und Experten der Entsorgungswirtschaft, der Glasindustrie und des glasverarbeitenden Gewerbes teil. Die mangelhafte Getrennthaltung der Flachglasscherben auf Baustellen von Rückbau- und Abrissprojekten führt zu unerwünschten Verunreinigungen der Scherben durch Erde und Steine. Daher besteht sowohl für die Steigerung der Quantität als auch der Qualität der Scherben noch einiges Potenzial. Als Ergebnis werden mehrere Maßnahmen genannt um dieses Ziel zu erreichen. Neben dem Recycling wurde auch die Wiederverwendung und die Vorbereitung zur Wiederverwendung von ausgebauten Fenstern diskutiert
A form-finding method for adaptive truss structures subject to multiple static load cases
No full textForm-finding is an essential task in the design of efficient lightweight structures. It is based on the crucial assumption of one single shape-determining load case, usually represented by self-weight. Adaptive components integrated into the structure open a way to even more efficient lightweight designs, as such structures can adapt their shapes to varying external loads and redistribute internal forces. This article presents a method for form-finding of adaptive truss structures subject to multiple, independently acting load cases, also incorporating possible design constraints. To ensure the consistency of the manufacturing lengths of passive elements in all load cases, special constraints are considered. The method enables to reduce sensitivity of the structural shape with respect to various different loads by means of actuation to meet design and serviceability requirements with a lower structural mass compared to conventional design strategies. This is demonstrated within a replaced real-world-like setting of an adaptive suspension truss bridge
Estimating Conditional Distributions with Neural Networks Using R Package deeptrafo
No full textContemporary empirical applications frequently require flexible regression models for complex response types and large tabular or non-tabular, including image or text, data. Classical regression models either break down under the computational load of processing such data or require additional manual feature extraction to make these problems tractable. Here, we present deeptrafo, a package for fitting flexible regression models for conditional distributions using a tensorflow back end with numerous additional processors, such as neural networks, penalties, and smoothing splines. Package deeptrafo implements deep conditional transformation models (DCTMs) for binary, ordinal, count, survival, continuous, and time series responses, potentially with uninformative censoring. Unlike other available methods, DCTMs do not assume a parametric family of distributions for the response. Further, the data analyst may trade off interpretability and flexibility by supplying custom neural network architectures and smoothers for each term in an intuitive formula interface. We demonstrate how to set up, fit, and work with DCTMs for several response types. We further showcase how to construct ensembles of these models, evaluate models using inbuilt cross-validation, and use other convenience functions for DCTMs in several applications. Lastly, we discuss DCTMs in light of other approaches to regression with non-tabular data
International Collaboration: Mainstreaming Artificial Intelligence and Cyberphysical Systems for Carbon Neutrality
No full textCyberphysical systems together with Artificial Intelligence play vital roles in reducing, eliminating, and removing greenhouse gas emissions across sectors. Electrification with renewables introduces complexity in systems in the deployment, integration, and efficient orchestration of electrified economic systems. AI-driven cyberphysical systems are uniquely suited to tackle this complexity, potentially accelerating the transition towards a low-carbon economy. The objective of this policy brief is to advocate for the mainstreaming of AI-driven cyberphysical systems for climate change risk mitigation and adaptation. To effectively and more rapidly realize the Intelligent Decarbonation potential, the concept of AI-driven cyberphysical systems must be elevated to a global level of collaboration and coordination, fostering research and development, capacity building, as well as knowledge and technology transfer. Drawing on a multidisciplinary, international study about intelligent decarbonization use cases, this brief also highlights factors impeding the transition to carbon neutrality and risks associated with technology determinism. The importance of governance is emphasized to avoid unwanted path dependency and avert a technology-solutionist approach dominating climate policy that delivers limited results. Given only 12% of the Sustainable Development Goals have been realized, a condensed version of this policy brief was submitted to the India T20, a G20 engagement group, urging global collaboration to prioritize AI-driven CPSs
Die Auswirkung von Managemententscheidungen: Wie Fachbereichs- und IT-Verantwortliche ihre IT-Kosten effektiv steuern können
Get PDFIT-Kosten machen heute einen immer größeren Anteil an den Gesamtkosten von Unternehmen aus. Die Verantwortlichen sind aufgefordert die IT-Kosten zu senken oder zumindest ein effizientes Management sicherzustellen. Oftmals fehlt es dafür an Transparenz und Verständnis für diese Ausgaben. Die Analyse der IT-Kostentreiber ermöglicht ein tieferes Verständnis der Ursachen und Auswirkungen strategischer Entscheidungen. Dieser Beitrag zielt darauf ab, die strategischen IT-Kostentreiber bezüglich des Wirkungshorizonts und des Entscheidungsortes zu analysieren. Die durchgeführte Delphi-Studie zeigt, dass Entscheidungen über diese Kostentreiber größtenteils mittel- bis langfristige Auswirkungen haben. Zudem wird deutlich, dass die IT-Abteilung zwar in den Entscheidungsprozess eingebunden ist, während die finalen Entscheidungen häufig stärker im Fachbereich liegen. Zusammenarbeit und effektive Kommunikation sind deshalb entscheidend und die Verantwortung für IT-Kosten sollte von allen EntscheidungsträgerInnen getragen werden. Dieser Beitrag erweitert die Forschung im IT-Kostenmanagement und sensibilisiert PraktikerInnen für Kostenbeeinflussungshebel und die strategische Diskussion über IT-Kosten und das Wertversprechen der IT
Bernstein flows for flexible posteriors in variational Bayes
Get PDFBlack-box variational inference (BBVI) is a technique to approximate the posterior of Bayesian models by optimization. Similar to MCMC, the user only needs to specify the model; then, the inference procedure is done automatically. In contrast to MCMC, BBVI scales to many observations, is faster for some applications, and can take advantage of highly optimized deep learning frameworks since it can be formulated as a minimization task. In the case of complex posteriors, however, other state-of-the-art BBVI approaches often yield unsatisfactory posterior approximations. This paper presents Bernstein flow variational inference (BF-VI), a robust and easy-to-use method flexible enough to approximate complex multivariate posteriors. BF-VI combines ideas from normalizing flows and Bernstein polynomial-based transformation models. In benchmark experiments, we compare BF-VI solutions with exact posteriors, MCMC solutions, and state-of-the-art BBVI methods, including normalizing flow-based BBVI. We show for low-dimensional models that BF-VI accurately approximates the true posterior; in higher-dimensional models, BF-VI compares favorably against other BBVI methods. Further, using BF-VI, we develop a Bayesian model for the semi-structured melanoma challenge data, combining a CNN model part for image data with an interpretable model part for tabular data, and demonstrate, for the first time, the use of BBVI in semi-structured models
Rechtliche Herausforderungen im Start-up-Marketing
No full textDeutschland erlebt zurzeit einen Boom an Unternehmensgründungen, vielfach KI-geprägt. Die Bandbreite an verschiedenen Arten von Start-ups ist dabei äußerst groß. Ziel dieses Buches ist es, die für Start-ups Verantwortlichen für die rechtlichen Aspekte ihrer Tätigkeit zu sensibilisieren. Es orientiert sich bei der Vorstellung der rechtlichen Rahmenbedingungen am Ablauf des Gründungsprozesses eines Start-ups, unterteilt in vorbereitende Maßnahmen, wie etwa die Anmeldung eines Gewerbebetriebes, den Schutz der Geschäftsidee, die Wahl der Rechtsform oder den Verträgen mit Geldgebern. Der zweite Abschnitt des Buches behandelt die eigentlichen Marketingmaßnahmen
Advanced Classifiers and Feature Reduction for Accurate Insomnia Detection Using Multimodal Dataset
No full textSleep deprivation is a significant contributor to various diseases, leading to poor cognitive function, decreased performance, and heart disorders. Insomnia, the most prevalent sleep disorder, requires more effective diagnosis and screening for proper treatment. Actigraphic data and its combination with physiological sensors like electroencephalogram (EEG), electrocardiogram (ECG), and body temperature have proven significant in predicting insomnia using machine learning methods. Studies focusing solely on actigraphic data achieved an accuracy of 84%, combining it with other wearable devices increased accuracy to 88%, and 2-channel EEG alone yielded an accuracy of 92%, but limits scalability and practicality in real-world settings. Here we show that using the hybrid approach of incorporating both recursive feature elimination (RFE) and principal component analysis (PCA) on sleep and heart data features yields outstanding results, with the multi-layer perception (MLP) achieving an accuracy of 95.83% and an F1 score of 0.93. The top-ranked features are predominantly sleep-related and time-domain RR interval. The dependent variables in our study have been extracted from the self-report Pittsburgh Sleep Quality Index questionnaire responses. Our findings emphasize the importance of tailoring feature sets and employing appropriate reduction techniques for optimal predictive modeling in sleep-related studies. Our results demonstrate that the ensemble classifiers generalize well on the dataset regardless of the feature count, while other algorithms are hindered by the curse of dimensionality