20005 research outputs found
Sort by
Guaranteed efficient energy estimation of quantum many-body Hamiltonians using ShadowGrouping
No full textEstimation of the energy of quantum many-body systems is a paradigmatic task in various research fields. In particular, efficient energy estimation may be crucial in achieving a quantum advantage for a practically relevant problem. For instance, the measurement effort poses a critical bottleneck for variational quantum algorithms. We aim to find the optimal strategy with single-qubit measurements that yields the highest provable accuracy given a total measurement budget. As a central tool, we establish tail bounds for empirical estimators of the energy. They are helpful for identifying measurement settings that improve the energy estimate the most. This task constitutes an NP-hard problem. However, we are able to circumvent this bottleneck and use the tail bounds to develop a practical, efficient estimation strategy, which we call ShadowGrouping. As the name indicates, it combines shadow estimation methods with grouping strategies for Pauli strings. In numerical experiments, we demonstrate that ShadowGrouping improves upon state-of-the-art meth-ods in estimating the electronic ground-state energies of various small molecules, both in provable and practical accuracy benchmarks. Hence, this work provides a promising way, e.g., to tackle the measurement bottleneck associated with quantum many-body Hamiltonians
ProMark : Ensuring Transparency and Privacy-Awareness in Proximity Marketing Advertising Campaigns
No full textAdvertising campaigns are crucial in business development, but most marketing techniques target online purchases (e.g., Google Adsense) and rely on a centralized architecture to store and process the campaign's data and check its effectiveness. Recently, proximity marketing has become more popular thanks to the widespread use of smartphones. It exploits the short-range communication (e.g., Bluetooth) between smartphones and beacon devices to collect and send marketing information to customers. However, this might create privacy issues for customers due to the potential leakage of sensitive information (such as locations associated with time). In this paper, we propose ProMark, a privacy-aware blockchain-based platform to verify the effectiveness of proximity marketing campaigns by ensuring transparency, decentralization, and privacy in the measurement process. We implemented ProMark and carried out experiments that show that ProMark can be used in super-regional malls even during peak hours
Robust design heuristics for product costing systems: a replication and extension using an ABC cost hierarchy
No full textAccurately reported product costs are essential for many managerial decisions, such as product and resource planning, pricing, product mix decisions, and cost management. Balakrishnan et al. (Manag Sci 57(3):520–541, 2011) offered valuable managerial guidance on how costing system design heuristics affect the accuracy of reported product costs. This paper tests the internal and external validity of their numerical experiments and recommendations. Our replication results show that we are able to reproduce most of the findings, thereby confirming the results’ internal validity. To test external validity and assess the robustness of their findings, we modify a key model element by implementing a resource consumption pattern that follows a four-tier Activity-Based Costing (ABC) cost hierarchy, with which we repeat the numerical experiments. Although the design heuristics are mostly robust in this modified environment, the benefits of improving the costing system design are less straightforward and less linear. For instance, single plant-wide cost drivers outperform more information-demanding costing systems with few (2–4) cost pools. Consequently, given a four-tier ABC cost hierarchy, refining the costing system incrementally can reduce accuracy, leaving costing system designers stuck in the middle. Overall, our study supports Balakrishnan et al. (2011) results’ robustness but also identifies reasons why firms may still use simple costing systems
Large language models for human-machine collaborative particle accelerator tuning through natural language
No full textAutonomous tuning of particle accelerators is an active and challenging research field with the goal of enabling advanced accelerator technologies and cutting-edge high-impact applications, such as physics discovery, cancer research, and material sciences. A challenge with autonomous accelerator tuning remains that the most capable algorithms require experts in optimization and machine learning to implement them for every new tuning task. Here, we propose the use of large language models (LLMs) to tune particle accelerators. We demonstrate on a proof-of-principle example the ability of LLMs to tune an accelerator subsystem based on only a natural language prompt from the operator, and compare their performance to state-of-the-art optimization algorithms, such as Bayesian optimization and reinforcement learning-trained optimization. In doing so, we also show how LLMs can perform numerical optimization of a nonlinear real-world objective. Ultimately, this work represents another complex task that LLMs can solve and promises to help accelerate the deployment of autonomous tuning algorithms to day-to-day particle accelerator operations
Explainable machine learning-based fatigue assessment of 316L stainless steel fabricated by laser-powder bed fusion
No full textAdditive manufacturing (AM) and in particular laser-powder bed fusion has become a popular manufacturing techniques in recent years due to its significant advantages; however, the mechanical behavior of AM components often varies from components fabricated using conventional processes. For example, the fatigue behavior of components made by AM processes is heavily influenced by process-related defects and residual stresses in addition to applied stress amplitudes, stress ratio and surface conditions. Accounting for the interaction of these effects in fatigue design is difficult by means of traditional fatigue assessment concepts. Machine learning algorithms offer a possibility to account for such interactions and are easily applied once trained and validated. In this study, machine learning algorithms based on gradient boosted trees with the SHapley Additive exPlanation framework are used to predict defect location and fatigue life of additive manufactured AISI 316L specimens in as-built and post-treated manufacturing states, while also facilitating the understanding of the importance and interactions of various influencing factors
Investigation on curing strategies for metal binder jetting with Ti-6Al-4V
No full textMetal binder jetting is a promising manufacturing technology that holds the potential to be a future competition technology to classic laser based additive manufacturing processes. In contrast to these technologies, however, metal binder jetting is much less mature. While sintering and debinding are already well known due to the spread of metal injection molding and powder deposition by laser powder bed fusion and its related processes, the often-neglected curing step represents a major challenge in process control. This study was therefore the first comprehensive investigation into the curing of metal binder jetting green parts from Ti-6Al-4 V powder with a powder size distribution below 25 µm. It was shown that the curing step has only a minor effect on the green part quality (surface roughness and density), but at the same time has a decisive influence on the green strength. In addition, position-dependent effects for the green density were detected, which indicate insufficient curing in the outer areas of the print box
A novel framework for estimating the Bowen ratio over small water bodies
No full textThe Bowen ratio, defined as the ratio of sensible to latent heat flux, is crucial for quantifying land-atmosphere energy exchanges and evaporation rates from terrestrial surfaces. Despite extensive research on the Bowen ratio over placid water surfaces (e.g., lakes), further investigation is needed to understand its dynamics in small reservoirs subjected to water inflow/outflow (i.e., surface flows) and wind. To address this knowledge gap, the evaporation rate and the sensible heat exchanges are measured between the water surface and overlying air in a small laboratory basin under different water surface flow rates (1.0–10.5 l min−1) and wind speeds (0–2.0 m s−1). Three different wind flow conditions are explored: no wind, headwind (opposing the water surface flow), and tailwind (aligning with water surface flow). The findings indicate strong correlations between sensible heat flux, water surface flow rate, and wind speed, particularly under headwind conditions. Nevertheless, concerning the latent heat flux, the measurements demonstrate that for each wind condition, the evaporation reaches its minimum value in a certain water surface flow rate, resulting in the highest value of the Bowen ratio. To facilitate the application of these laboratory findings for estimating the Bowen ratio under real environmental conditions, mathematical relationships using dimensionless numbers obtained through non-linear regression analysis are established. The results exhibit a good agreement with measurements in a small water basin
A novel, cost-effective approach for the production of hydrogenase enzymes and molecular hydrogen from recycled whey-based by-products
No full textCupriavidus necator produces O2-tolerant [NiFe]-hydrogenases (Hyds), and is a valuable biocatalyst for biological fuel cells, while Escherichia coli produces H2 during mixed-acid fermentation. A mixture of curd and cheese whey (CW) was used to explore two-phase growth with C. necator H16 and E. coli. Enhanced biomass, as well as Hyd activity and electrical potential (∼0.65 V) were shown for growth of C. necator in the CW medium with added glycerol. The residual growth medium made available after the cultivation of C. necator and removal of cells was then used for cultivation of E. coli. Maximum fermentative growth of E. coli was attained after 72 h and with a H2 yield of ∼6 mmol/L/g dry whey after 48 h. This study demonstrates the economically viable production of biomass, hydrogenase enzymes and H₂ using cheap, industrially produced whey within the 3R concept
Structural benefits on prescription? Exploring potentials and challenges of the pSVV concept in the approval process of digital health applications in Germany
No full textHintergrund
Mit dem Inkrafttreten des Digitale-Versorgung-Gesetzes (DVG) Ende 2019 wurden in Deutschland Digitale Gesundheitsanwendungen (DiGA) in den Leistungskatalog der gesetzlichen Krankenversicherung (GKV) aufgenommen. DiGA sind digitale Medizinprodukte, deren Hauptfunktionen wesentlich auf digitalen Technologien beruhen. Im Rahmen der Definition des Zulassungsprozesses wurde der in der Arzneimittelzulassung zur Nutzenbewertung maßgebliche Begriff des „therapeutischen Nutzens“ zum umfassenderen Konzept des „positiven Versorgungseffekts“ (pVE) erweitert. Konkret wurde ergänzend zum „medizinischen Nutzen“ das Konzept der „patientenrelevanten Struktur- und Verfahrensverbesserungen“ (pSVV) entwickelt, um den Markteintritt von Anwendungen zu ermöglichen, die die Rolle von Patient*innen in der Gesundheitsversorgung gezielt stärken. Dreieinhalb Jahre nach der Einführung von DiGA zeichnet sich jedoch ab, dass das Konzept des pSVV heute noch nicht von den Akteur*innen im Gesundheitssystem akzeptiert wurde. Lediglich eine der 56 zum 1. Juli 2024 gelisteten DiGA nutzt heute pSVV als primären Endpunkt, zehn weitere DiGA nutzen pSVV als sekundären Endpunkt.
Methode
Ein qualitativer Ansatz wurde gewählt, um die neuen und wenig erforschten Themen DiGA und insbesondere pSVV zu untersuchen. Dabei wurde die Grounded-Theory-Methode in Kombination mit der Gioia-Methode angewendet, die sich besonders für die Analyse innovativer Themengebiete eignet. Durch die induktive Herangehensweise können so neue Konzepte aus den Daten der Untersuchungsteilnehmer*innen entwickelt werden, wodurch eine flexible und dynamische Theoriebildung unterstützt wird. Es wurden Entscheidungsträger*innen aus den Gruppen DiGA-Herstellende mit und ohne pSVV, Herstellende digitaler Medizinprodukte ohne DiGA-Zulassung, Beratungen sowie am Prozess der DiGA-Zulassung beteiligte Institutionen mithilfe der Durchführung qualitativer Leitfadeninterviews in die Untersuchung einbezogen. Die Auswertung erfolgte durch eine mehrstufige Analyse, die zur Identifikation von Erstordnungskonzepten, Zweitordnungsthemen und aggregierten Dimensionen führte. Die entwickelte Datenstruktur wurde durch die Darstellung relevanter Zitate im Anhang A untermauert.
Ergebnisse
Es zeigt sich, dass eine Schärfung der Nutzendefinition der im DiGA-Leitfaden definierten pSVV-Kategorien erfolgskritisch für die Etablierung des Konzepts der pSVV ist. Auch eine weitere Analyse der möglichen Methoden zum Nachweis des pSVV und geeigneter Messinstrumente sowie die generelle Aufnahme ökonomischer Betrachtungen in die Nutzenanalyse erscheinen sinnvoll.
Diskussion
Die qualitative Analyse zeigt, dass die Schärfung des pSVV-Konzepts und die Definition geeigneter Messmethoden entscheidend für die erfolgreiche Implementierung im deutschen Gesundheitssystem sind. Eine ökonomische Bewertung könnte die Debatte um die Kosten von DiGA versachlichen und zur Transparenz im Zulassungsprozess beitragen. Weitere Forschung und die Einbindung zentraler Akteure sind notwendig, um die Definition des intendierten Nutzens innerhalb der pSVV-Kategorien zu präzisieren und damit die Einführung digitaler Gesundheitsanwendungen mit besonderem Fokus auf die Stärkung der Rolle der Patient*innen in der Gesundheitsversorgung zu fördern