Publikationsserver der Westfälischen Hochschule
Not a member yet
2819 research outputs found
Sort by
Also in Chemistry, Deep Learning Models Love Really Big Data
No full textInspired by the super-human performance of deep learning models in playing the game of Go after being presented with virtually unlimited training data, we looked into areas in chemistry where similar situations could be achieved. Encountering large amounts of training data in chemistry is still rare, so we turned to two areas where realistic training data can be fabricated in large quantities, namely a) the recognition of machine-readable structures from images of chemical diagrams and b) the conversion of IUPAC(-like) names into structures and vice versa. In this talk, we outline the challenges, technical implementation and results of this study.
Optical Chemical Structure Recognition (OCSR): Vast amounts of chemical information remain hidden in the primary literature and have yet to be curated into open-access databases. To automate the process of extracting chemical structures from scientific papers, we developed the DECIMER.ai project. This open-source platform provides an integrated solution for identifying, segmenting, and recognising chemical structure depictions in scientific literature. DECIMER.ai comprises three main components: DECIMER-Segmentation, which utilises a Mask-RCNN model to detect and segment images of chemical structure depictions; DECIMER-Image Classifier EfficientNet-based classification model identifies which images contain chemical structures and DECIMER-Image Transformer which acts as an OCSR engine which combines an encoder-decoder model to convert the segmented chemical structure images into machine-readable formats, like the SMILES string.
DECIMER.ai is data-driven, relying solely on the training data to make accurate predictions without hand-coded rules or assumptions. The latest model was trained with 127 million structures and 483 million depictions (4 different per structure) on Google TPU-V4 VMs
Name to Structure Conversion: The conversion of structures to IUPAC(-like) or systematic names has been solved algorithmically or rule-based in satisfying ways. This fact, on the other side, provided us with an opportunity to generate a name-structure training pair at a very large scale to train a proof-of-concept transformer network and evaluate its performance.
In this work, the largest model was trained using almost one billion SMILES strings. The Lexichem software utility from OpenEye was employed to generate the IUPAC names used in the training process. STOUT V2 was trained on Google TPU-V4 VMs. The model's accuracy was validated through one-to-one string matching, BLEU scores, and Tanimoto similarity calculations. To further verify the model's reliability, every IUPAC name generated by STOUT V2 was analysed for accuracy and retranslated using OPSIN, a widely used open-source software for converting IUPAC names to SMILES. This additional validation step confirmed the high fidelity of STOUT V2's translations
Johann Friedrich Pfeiffer on Adam Smith: An Early Reception of Adam Smith in the German States
No full textDifferent from the usual portrayal that it was not until the turn of the nineteenth century that Adam Smith’s work became influential, the extensive commentary by Johann Friedrich Pfeiffer shows an immediate effect on German economic discourse. Pfeiffer, who lived from 1717 until 1787, was as a late and liberal cameralist. Pfeiffer was a prolific writer and well-known scholar during his lifetime. He lived so close to 1789 that his works were only briefly received, and he was not able to leave a lasting legacy on cameralist thought, which withered after the French Revolution. The significance of Smith was recognized by Pfeiffer, placing Smith above most of his German-speaking contemporaries. Both Pfeiffer and Smith addressed many similar topics and Pfeiffer expresses his agreement with large parts of Smith’s Wealth of Nations. However, Pfeiffer criticizes Smith because of Smith’s general support of free trade and his idealistic concept of a system of natural liberty. Pfeiffer, in contrast, was much more in favor of state interventions, given the lack of knowledge and irrational behavior of humans. While their political points of view differ, there are many theoretical similarities between them
Analysis of the Potential for Thermal Flexibility of Cooling Applications
No full textThe feed-in of electricity from renewable energies, such as wind or solar power, fluctuates based on weather conditions. This unpredictability due to volatile feed-in can lead to sudden changes in energy generation so that solutions ensuring grid stability need to be implemented. The cooling sector offers the opportunity to create flexibilities for such balancing, with this study focusing on the thermal flexibilities that can be provided by cooling applications. Various cooling-demand profiles are investigated with respect to their load profile and their impact on flexibility is analysed. In addition to the cooling demand, scenarios of different storage dimensions are considered. As a result, it shows that an increasing base-load level and increasing operating-load duration have a negative effect on flexibility, while an increasing full-load duration is beneficial for flexibility. Storage size also has a strong impact as higher storage capacity and storage performance indicate higher flexibility, whereas above a certain size they only provide little added value
Thermal 360° micro drone: Operational exercise under smoke at the Dortmund fire house
No full textThe 360° video shows an operational exercise under smoke of our new 360° + thermal camera video mini drone (18x18x18 cm) in the fire house in Dortmund. The thermal video is overlaid with the 360° video. The exercise was also accompanied by WDR (https://www.ardmediathek.de/video/wdr-dok/unser-leben-mit-ki-wie-kuenstliche-intelligenz-unsere-arbeit-revolutioniert/wdr/Y3JpZDovL3dkci5kZS9CZWl0cmFnLXNvcGhvcmEtODRjYWI5NjQtYjAxYS00NjdiLThjODgtYzViMGVmNTY3OThj from minute 7:16). The winds occurring during the fire are a particular challenge for the small drone and the pilot
Unterstützung wirksamer Führung durch personaldiagnostisch gestütztes (Business-)Coaching
No full textZusammenfassung
Im vorliegenden Beitrag geht es um die Frage, wie personaldiagnostisch gestütztes (Business-)Coaching zur Entwicklung von Führungskompetenzen und Unterstützung wirksamer Führung eingesetzt werden kann. Nach einer kurzen Einführung in aktuelle Herausforderungen und zukünftige Anforderungen an wirksame Führung werden zunächst Grundlagen der Potenzialentwicklung von Führungskräften durch Coaching diskutiert. Neben der Definition und Abgrenzung von Coaching werden hierzu Coachinganlässe und daraus resultierende Anforderungen an die Coaches betrachtet, Wirkmechanismen im Coaching aufgezeigt und grundlegende Einsatzfelder und -möglichkeiten personaldiagnostischer Instrumente im Coaching erläutert. Als Anwendungsbeispiel aus der Praxis wird anschließend der RAUEN Analyzer® als personaldiagnostisches Instrument im Führungskräftecoaching vorgestellt. Neben der wissenschaftlichen Fundierung des Instruments werden exemplarische Ansatzpunkte, die das Instrument zur Unterstützung wirksamer Führung liefert, an zwei Fallbeispielen dargestellt und diskutiert und so Möglichkeiten, aber auch Grenzen eines personaldiagnostisch gestützten Coachings im Führungskontext aufgezeigt
Why does the robot only select men? How women and men perceive autonomous social robots that have a gender bias
No full textAbstract
Future social robots will act autonomously in the world. Autonomous behavior is usually realized by using AI models built with real-world data, which often reflect existing inequalities and prejudices in society. Even if designers do not intend it, there are risks that robots will be developed that discriminate against certain users, e. g. based on gender. In this work, we investigate the implications of a gender-biased robot that disadvantages women, which unfortunately is a bias in AI that is often reported. Our experiment shows that both men and women perceive the gender-biased robot to be unfair. However, our work indicates that women are more aware that a gender bias causes this unfairness. We also show that gender bias results in the robot being perceived differently. While the gender bias resulted in lower likability and intelligence ratings by women, men seem to lose trust in the robot if it behaves unfairly
Manuelle Programmierung eines 3D Druckers
Get PDFDie Anwendung des "FullControll GCode Designer" vereinfacht den 3D-Druckprozess, indem er die 3D-Modellierung und den Einsatz eines Slicer-Programms überspringt und stattdessen direkt den G-Code erstellt. Die vorgefertigte Excel-Anwendung ermöglicht es, Objekte durch Angabe der Start- und Zielkoordinaten effizient Linie für Linie mit minimalem Eingabeaufwand zu programmieren, wobei verschiedene Druckparameter angepasst werden können, um unterschiedliche Effekte zu erzielen. In diesem Werk werden die Möglichkeiten und Grenzen des Designers erarbeitet
Gesamtpreis im Sinne der PAngV enthält nicht den beim Kauf von Waren in Pfandbehältern zu entrichtenden Pfandbetrag (BGH 26.10.2023 - I ZR 135/20)
No full textSTOUT V2.0: SMILES to IUPAC name conversion using transformer models
No full textNaming chemical compounds systematically is a complex task governed by a set of rules established by the International Union of Pure and Applied Chemistry (IUPAC). These rules are universal and widely accepted by chemists worldwide, but their complexity makes it challenging for individuals to consistently apply them accurately. A translation method can be employed to address this challenge. Accurate translation of chemical compounds from SMILES notation into their corresponding IUPAC names is crucial, as it can significantly streamline the laborious process of naming chemical structures. Here, we present STOUT (SMILES-TO-IUPAC-name translator) V2, which addresses this challenge by introducing a transformer-based model that translates string representations of chemical structures into IUPAC names. Trained on a dataset of nearly 1 billion SMILES strings and their corresponding IUPAC names, STOUT V2 demonstrates exceptional accuracy in generating IUPAC names, even for complex chemical structures. The model’s ability to capture intricate patterns and relationships within chemical structures enables it to generate precise and standardised IUPAC names. While established deterministic algorithms remain the gold standard for systematic chemical naming, our work, enabled by access to OpenEye’s Lexichem software through an academic license, demonstrates the potential of neural approaches to complement existing tools in chemical nomenclature