Showing papers by "Johannes Kepler University of Linz published in 2022"

Hydrological concept formation inside long short-term memory (LSTM) networks

Abstract: Abstract. Neural networks have been shown to be extremely effective rainfall-runoff models, where the river discharge is predicted from meteorological inputs. However, the question remains: what have these models learned? Is it possible to extract information about the learned relationships that map inputs to outputs, and do these mappings represent known hydrological concepts? Small-scale experiments have demonstrated that the internal states of long short-term memory networks (LSTMs), a particular neural network architecture predisposed to hydrological modelling, can be interpreted. By extracting the tensors which represent the learned translation from inputs (precipitation, temperature, and potential evapotranspiration) to outputs (discharge), this research seeks to understand what information the LSTM captures about the hydrological system. We assess the hypothesis that the LSTM replicates real-world processes and that we can extract information about these processes from the internal states of the LSTM. We examine the cell-state vector, which represents the memory of the LSTM, and explore the ways in which the LSTM learns to reproduce stores of water, such as soil moisture and snow cover. We use a simple regression approach to map the LSTM state vector to our target stores (soil moisture and snow). Good correlations (R2>0.8) between the probe outputs and the target variables of interest provide evidence that the LSTM contains information that reflects known hydrological processes comparable with the concept of variable-capacity soil moisture stores. The implications of this study are threefold: (1) LSTMs reproduce known hydrological processes. (2) While conceptual models have theoretical assumptions embedded in the model a priori, the LSTM derives these from the data. These learned representations are interpretable by scientists. (3) LSTMs can be used to gain an estimate of intermediate stores of water such as soil moisture. While machine learning interpretability is still a nascent field and our approach reflects a simple technique for exploring what the model has learned, the results are robust to different initial conditions and to a variety of benchmarking experiments. We therefore argue that deep learning approaches can be used to advance our scientific goals as well as our predictive goals.

Accelerating sustainable and economic development via industrial energy cooperation and shared services – A case study for three European countries

Abstract: The European industry sector is responsible for about one-third of the EU's total energy consumption and process-related Greenhouse Gas emissions, which makes it a central factor in the EU's climate and energy strategies. Energy cooperation, i.e., the mutualised generation, use and/or acquisition of energy by at least two companies, has the potential of significantly supporting the successful achievement of these strategies. This paper presents a concept for moving from a single-company sustainable energy intervention approach to cooperative sustainable energy solutions within the framework of industrial parks. Technical, economic, regulatory, organizational, and social barriers for energy-efficient park design and operation on all levels and instruments to overcome them have been systematically analysed, taking correlation of solutions, and park-specific requirements into account, thus providing a holistic workflow. First results show that technical and economic attractiveness, and an enabling legal and policy context, are not always enough. For promising energy cooperation solutions to flourish, cultural, organizational, social, and behavioural factors also play a significant role. Furthermore, findings show that external facilitators are helpful to pool efforts to achieve greater engagement.

A Trajectory-Based Approach to Discrete-Time Flatness

Abstract: For discrete-time systems, flatness is usually defined by replacing the time-derivatives of the well-known continuous-time definition by forward-shifts. With this definition, the class of flat systems corresponds exactly to the class of systems which can be linearized by a discrete-time endogenous dynamic feedback as it is proposed in the literature. Recently, verifiable necessary and sufficient differential-geometric conditions for this property have been derived. In the present contribution, we make an attempt to take into account also backward-shifts. This extended approach is motivated by the one-to-one correspondence of solutions of flat systems to solutions of a trivial system as it is known from the continuous-time case. If we transfer this idea to the discrete-time case, this leads to an approach which also allows backward-shifts. To distinguish the classical definition with forward-shifts and the approach of this letter, we refer to the former as forward-flatness. We show that flat systems (in the extended sense with backward-shifts) still share many beneficial properties of forward-flat systems. In particular, they still are reachable/controllable, allow a straightforward planning of trajectories and can be linearized by a certain subclass of dynamic feedbacks.

Valvuloplasty in 103 fetuses with critical aortic stenosis: outcome and new predictors for postnatal circulation

Abstract: Objectives Patients with prenatally diagnosed critical aortic stenosis (CAS) and evolving hypoplastic left heart syndrome (eHLHS) still carry significant morbidity and mortality. The aims of this study were to review our 20-year experience with fetal aortic valvuloplasty (FAV) in fetuses with CAS and eHLHS including early and mid-term postnatal outcome and to refine selection criteria for FAV. Methods This was a retrospective review of all fetuses with CAS and eHLHS undergoing FAV at our center since 2001. Echocardiograms and patient charts were analyzed for ventricular and valvular dimensions, as well as for hemodynamics and outcome. CART analysis was performed to investigate the predictive capacities of analyzed parameters. Results Since 2001 103 fetuses underwent 125 FAV at our center. 87.4% of fetuses had a technical successful procedure. Technical success per fetus was higher in the recent period since 2014 (96.2% (51/53) vs. 78%, p= 0.0068). 78 fetuses were liveborn after successful intervention and received further treatment. Biventricular outcome at one year of age could be achieved in 55% of liveborn patients and was significantly better when compared to those patients of the natural history cohort (23.7%) with similar inclusion criteria (p= 0.0015). Decision tree analysis based on the ratios of right and left ventricular length combined with LV pressure estimates (mitral regurgitation Vmax) showed a high sensitivity and specificity of 96.97% and 94.44% with regards to biventricular outcome without signs of PAH at one year of age. Conclusion FAV could be performed with high success rates and an acceptable risk with improving results after a learning curve. Ratios of right and left ventricular length combined with LV pressure estimates were able to predict a successful BV outcome with high sensitivity and specificity. This article is protected by copyright. All rights reserved.

A Survey of Data Quality Measurement and Monitoring Tools

Abstract: High-quality data is key to interpretable and trustworthy data analytics and the basis for meaningful data-driven decisions. In practical scenarios, data quality is typically associated with data preprocessing, profiling, and cleansing for subsequent tasks like data integration or data analytics. However, from a scientific perspective, a lot of research has been published about the measurement (i.e., the detection) of data quality issues and different generally applicable data quality dimensions and metrics have been discussed. In this work, we close the gap between data quality research and practical implementations with a detailed investigation on how data quality measurement and monitoring concepts are implemented in state-of-the-art tools. For the first time and in contrast to all existing data quality tool surveys, we conducted a systematic search, in which we identified 667 software tools dedicated to "data quality." To evaluate the tools, we compiled a requirements catalog with three functionality areas: (1) data profiling, (2) data quality measurement in terms of metrics, and (3) automated data quality monitoring. Using a set of predefined exclusion criteria, we selected 13 tools (8 commercial and 5 open-source tools) that provide the investigated features and are not limited to a specific domain for detailed investigation. On the one hand, this survey allows a critical discussion of concepts that are widely accepted in research, but hardly implemented in any tool observed, for example, generally applicable data quality metrics. On the other hand, it reveals potential for functional enhancement of data quality tools and supports practitioners in the selection of appropriate tools for a given use case.

Wide-bandgap organic solar cells with a novel perylene-based non-fullerene acceptor enabling open-circuit voltages beyond 1.4 V

Abstract: A perylene-based acceptor (PMI-FF-PMI), consisting of two perylene monoimide (PMI) units bridged with a dihydroindeno[1,2-b]fluorene molecule was developed as a potential non-fullerene acceptor (NFA) for organic solar cells (OSCs). The synthesized NFA was combined with the high-performance donor polymer D18 to fabricate efficient OSCs. With an effective bandgap of 2.02 eV, the D18:PMI-FF-PMI blend can be categorized as a wide-bandgap OSC and is an attractive candidate for application as a wide-bandgap sub-cell in all-organic triple-junction solar cell devices. Owing to their large effective bandgap, D18:PMI-FF-PMI solar cells are characterized by an extremely high open-circuit voltage (VOC) of 1.41 V, which to the best of our knowledge is the highest reported value for solution-processed OSCs so far. Despite the exceptionally high VOC of this blend, a comparatively large non-radiative voltage loss (ΔVnon-radOC) of 0.25 V was derived from a detailed voltage loss analysis. Measurements of the electroluminescence quantum yield (ELQY) of the solar cell reveal high ELQY values of ∼0.1%, which contradicts the ELQY values derived from the non-radiative voltage loss (ΔVnon-radOC = 0.25 V, ELQY = 0.0063%). This work should help to raise awareness that (especially for BHJ blends with small ΔHOMO or ΔLUMO offsets) the measured ELQY cannot be straightforwardly used to calculate the ΔVnon-radOC. To avoid any misinterpretation of the non-radiative voltage losses, the presented ELQY discrepancies for the D18:PMI-FF-PMI system should encourage OPV researchers to primarily rely on the ΔVnon-radOC values derived from the presented voltage loss analysis based on EQEPV and J-V measurements.

Scattering <i>versus</i> fluorescence self-quenching: more than a question of faith for the quantification of water flux in large unilamellar vesicles?

Abstract: The endeavors to understand the determinants of water permeation through membrane channels, the effect of the lipid or polymer membrane on channel function, the development of specific water flow inhibitors, the design of artificial water channels and aquaporins for the use in industrial water filtration applications all rely on accurate ways to quantify water permeabilities (Pf). A commonly used method is to reconstitute membrane channels into large unilamellar vesicles (LUVs) and to subject these vesicles to an osmotic gradient in a stopped-flow device. Fast recordings of either scattered light intensity or fluorescence self-quenching signals are taken as a readout for vesicle volume change, which in turn can be recalculated to accurate Pf values. By means of computational and experimental data, we discuss the pros and cons of using scattering versus self-quenching experiments or subjecting vesicles to hypo- or hyperosmotic conditions. In addition, we explicate for the first time the influence of the LUVs size distribution, channel distribution between vesicles and remaining detergent after protein reconstitution on Pf values. We point out that results such as the single channel water permeability (pf) depend on the membrane matrix or on the direction of the applied osmotic gradient may be direct results of the measurement and analysis procedure.

The Hidden Intricacies of Aquaporins: Remarkable Details in a Common Structural Scaffold

Abstract: Evolution turned aquaporins (AQPs) into the most efficient facilitators of passive water flow through cell membranes at no expense of solute discrimination. In spite of a plethora of solved AQP structures, many structural details remain hidden. Here, by combining extensive sequence- and structural-based analysis of a unique set of 20 non-redundant high-resolution structures and molecular dynamics simulations of four representatives, key aspects of AQP stability, gating, selectivity, pore geometry, and oligomerization, with a potential impact on channel functionality, are identified. The general view of AQPs possessing a continuous open water pore is challenged and it is depicted that AQPs' selectivity is not exclusively shaped by pore-lining residues but also by the relative arrangement of transmembrane helices. Moreover, this analysis reveals that hydrophobic interactions constitute the main determinant of protein thermal stability. Finally, a numbering scheme of the conserved AQP scaffold is established, facilitating direct comparison of, for example, disease-causing mutations and prediction of potential structural consequences. Additionally, the results pave the way for the design of optimized AQP water channels to be utilized in biotechnological applications.

Zu Datengrundlage und Datenqualität: Methodische Reflexion zur quantitativen Erhebung während der Corona-Krise

Abstract: Zusammenfassung Ziel von sozialwissenschaftlichen Umfragen ist es inhaltlich relevante und für die jeweilige Population aussagekräftige Daten zu generieren. Die Corona-Krise hat die Umfrageforschung in mehrfacher Hinsicht vor Herausforderungen gestellt: 1) Es war aufgrund der Rahmenbedingungen während der Pandemie nicht möglich etablierte Erhebungsformen, wie das Face-to-Face-Interview, zu nutzen. 2) Aufgrund der spezifischen Dynamik der Krise war es ebenso unmöglich, längerfristig Erhebungsmethode, Datenerhebungsinstrumente und die Umsetzung der Studien zu planen. Das folgende Kapitel präsentiert mit dem Austrian Corona Panel Project und der österreichischen Values in Crisis Umfrage die zwei zentralen Erhebungen, auf denen der Sammelband Österreich in der Corona-Krise. Ergebnisse aus sozialwissenschaftlichen Umfragen aufbaut. Neben einer Vorstellung der Erhebungsdesigns werden überblicksartig Informationen zur Datenqualität der Erhebungen präsentiert.

Do Men Have No Need for “Feminist” Artificial Intelligence? Agentic and Gendered Voice Assistants in the Light of Basic Psychological Needs

Abstract: Artificial Intelligence (AI) is supposed to perform tasks autonomously, make competent decisions, and interact socially with people. From a psychological perspective, AI can thus be expected to impact users' three Basic Psychological Needs (BPNs), namely (i) autonomy, (ii) competence, and (iii) relatedness to others. While research highlights the fulfillment of these needs as central to human motivation and well-being, their role in the acceptance of AI applications has hitherto received little consideration. Addressing this research gap, our study examined the influence of BPN Satisfaction on Intention to Use (ITU) an AI assistant for personal banking. In a 2×2 factorial online experiment, 282 participants (154 males, 126 females, two non-binary participants) watched a video of an AI finance coach with a female or male synthetic voice that exhibited either high or low agency (i.e., capacity for self-control). In combination, these factors resulted either in AI assistants conforming to traditional gender stereotypes (e.g., low-agency female) or in non-conforming conditions (e.g., high-agency female). Although the experimental manipulations had no significant influence on participants' relatedness and competence satisfaction, a strong effect on autonomy satisfaction was found. As further analyses revealed, this effect was attributable only to male participants, who felt their autonomy need significantly more satisfied by the low-agency female assistant, consistent with stereotypical images of women, than by the high-agency female assistant. A significant indirect effects model showed that the greater autonomy satisfaction that men, unlike women, experienced from the low-agency female assistant led to higher ITU. The findings are discussed in terms of their practical relevance and the risk of reproducing traditional gender stereotypes through technology design.

The landscape of retesting in childhood-onset idiopathic growth hormone deficiency and its reversibility: a systematic review and meta-analysis

Abstract: Children diagnosed with idiopathic isolated growth hormone deficiency (IGHD) are frequently observed to no longer be GH-deficient at a later stage of growth as a result of 'GHD reversal'. Reevaluation of GH status by stimulation test is currently incorporated into management guidelines at attainment of final height (FH). Over the past three decades, numerous studies have evaluated reversal rates using different methodologies including crucial parameters like GHD aetiology, GH cut-off and retesting time point, with heterogeneous results. We aimed to systematically analyse the reversibility of childhood-onset IGHD dependent on retesting GH cut-offs and retesting time points.PubMed, Cochrane Library, TRIP database and NHS Evidence were searched for publications investigating the reversibility of IGHD from database initiation to 30 June 2020 following PRISMA recommendations. Study cohorts were pooled according to retesting GH cut-off and time point. Reversal rates were calculated using random-effects models.Of the 29 studies initially identified, 25 provided sufficient detail for IGHD analysis, resulting in 2030 IGHD patient data. Reversal rates decreased significantly as the retesting GH cut-off increased (P = 0.0013). Pooled (95% CI) reversal rates were 80% (59-92%, n = 227), 73% (62-81%, n = 516) and 55% (41-68%, n = 1287) for cohorts using retesting GH cut-offs of 3-4 ng/mL, 5-6 ng/mL and 7.7-10 ng/mL, respectively. Individuals retested at FH (n = 674) showed a pooled reversal rate of 74% (64-82%) compared to 48% (25-71%) when retested before FH (n = 653).Provided evidence supports reevaluation of current IGHD management guidelines. The high reversal rates should instigate consideration of early retesting.

Flatness-Based Output Feedback Tracking Control of a Hyperbolic Distributed-Parameter System

Abstract: The hyperbolic distributed-parameter system under consideration comprises a second-order partial differential equation that is bidirectionally coupled with a first-order ordinary differential equation. The system models the transmission at a pneumatic test bench with boundary control and collocated measurement, yet stands in place for several other lossy or loss-free transmission problems. In the paper, an observer-based output feedback tracking controller is designed using the concept of flatness. It is a design based on normal forms, as both the state feedback tracking controller and the state observer are derived using canonical coordinates, which in turn rely on special parametrizations of the system’s solutions. Additionally, a new transformation between the coordinates of the so-called hyperbolic controller and observer canonical form is presented that allows for the implementation of the observer-based controller. Overall, the closed-loop system is exponentially stable. Simulation results illustrate the tracking behavior.

Combined person classification with airborne optical sectioning

Abstract: Fully autonomous drones have been demonstrated to find lost or injured persons under strongly occluding forest canopy. Airborne optical sectioning (AOS), a novel synthetic aperture imaging technique, together with deep-learning-based classification enables high detection rates under realistic search-and-rescue conditions. We demonstrate that false detections can be significantly suppressed and true detections boosted by combining classifications from multiple AOS-rather than single-integral images. This improves classification rates especially in the presence of occlusion. To make this possible, we modified the AOS imaging process to support large overlaps between subsequent integrals, enabling real-time and on-board scanning and processing of groundspeeds up to 10 m/s.