Last 7 days
Bookmark and Share    
Full Text
Peer Reviewed
See detailFedFog: Network-Aware Optimization of Federated Learning over Wireless Fog-Cloud System
Nguyen, van Dinh UL; Chatzinotas, Symeon UL; Ottersten, Björn UL et al

in IEEE Transactions on Wireless Communications (in press)

Federated learning (FL) is capable of performing large distributed machine learning tasks across multiple edge users by periodically aggregating trained local parameters. To address key challenges of ... [more ▼]

Federated learning (FL) is capable of performing large distributed machine learning tasks across multiple edge users by periodically aggregating trained local parameters. To address key challenges of enabling FL over a wireless fogcloud system (e.g., non-i.i.d. data, users’ heterogeneity), we first propose an efficient FL algorithm based on Federated Averaging (called FedFog) to perform the local aggregation of gradient parameters at fog servers and global training update at the cloud. Next, we employ FedFog in wireless fog-cloud systems by investigating a novel network-aware FL optimization problem that strikes the balance between the global loss and completion time. An iterative algorithm is then developed to obtain a precise measurement of the system performance, which helps design an efficient stopping criteria to output an appropriate number of global rounds. To mitigate the straggler effect, we propose a flexible user aggregation strategy that trains fast users first to obtain a certain level of accuracy before allowing slow users to join the global training updates. Extensive numerical results using several real-world FL tasks are provided to verify the theoretical convergence of FedFog. We also show that the proposed co-design of FL and communication is essential to substantially improve resource utilization while achieving comparable accuracy of the learning model. [less ▲]

Detailed reference viewed: 63 (10 UL)
Full Text
Peer Reviewed
See detailEnergy performance comparison of ventilation-based heating versus water-based heating systems in an efficient residential building
Shirani, Arsalan UL; Merzkirch, Alexander; Leyer, Stephan UL et al

in Fluid Dynamics & Material Processing (in press)

The application of air-based heating systems as a possible approach to reduce the construction costs in highly efficient residential buildings is becoming popular. Air-based heating systems have been well ... [more ▼]

The application of air-based heating systems as a possible approach to reduce the construction costs in highly efficient residential buildings is becoming popular. Air-based heating systems have been well-known for their usage in passive houses during the past three decades. Available studies on such systems tend mostly to focus only on comparing exhaust air heat pump technology with conventional systems in efficient buildings. Moreover, most of the existing studies ignore the usual presence of the electrical heaters as backup. Besides, a comprehensive study and comparison between different air-based heating system concepts is still missing. In this study, four different air-based heating system concepts separated by the type of heat source of heat pump for heating and domestic hot water are defined. These systems are compared to four conventional heating system, including floor heating and direct electrical system employing dynamic annual simulations. According to simulation results, the systems with floor heating have shown the best system efficiencies and the lowest energy demand in comparison to the other systems. The main reason for this was the lower supply temperatures of the floor heating systems. Between the air heating systems, the system equipped with an outdoor air heat pump showed a better energy performance than an exhaust air system. The main reason for this could be attributed to the power limitation of exhaust air heat pump systems. [less ▲]

Detailed reference viewed: 45 (1 UL)
See detailProtéger, libérer, assujettir. L’extension territoriale de la commune de Florence au XIVe siècle
Abeles, Solal UL

Book published by École française de Rome (in press)

Detailed reference viewed: 19 (0 UL)
Full Text
Peer Reviewed
See detailSprachliche Mittel in Fake News: Eine textlinguistische Perspektive
Huemer, Birgit UL

in Bendheim, Amelie; Pavlik, Jennifer (Eds.) Fake News. Von Fakten und Fiktionen in Literatur und Medien (in press)

Linguistische Studien zum Thema Fake News nähern sich diesem Phänomen hauptsächlich aus zwei verschiedenen Richtungen. Mit Methoden der kritischen Diskursanalyse wird untersucht wie der Begriff Fake News ... [more ▼]

Linguistische Studien zum Thema Fake News nähern sich diesem Phänomen hauptsächlich aus zwei verschiedenen Richtungen. Mit Methoden der kritischen Diskursanalyse wird untersucht wie der Begriff Fake News im öffentlichen Diskurs verwendet wird, welche Vorstellungen geschaffen, welche Identitäten konstruiert werden und welche Machtverhältnisse erzeugt, verteidigt oder herausgefordert werden. Computerlinguistische Ansätze beschäftigen sich mit der Charakterisierung und automatisierten Erkennung von Fake News, um diese von sogenanntem „rechtmäßigen“ Nachrichten (legitimate news ) zu unterscheiden. Textlinguistische Untersuchungen zum Thema Fake News gibt es bisher kaum. In diesem Beitrag wird zuerst den wenig vorhandenen linguistischen Definitionsansätzen von Fake News auf den Grund gegangen, mit dem Ziel diese durch die textlinguistische Analyse zu ergänzen. Danach werden die beiden sprachwissenschaftlichen Zugänge zum Phänomen Fake News vorgestellt und einige Untersuchungen zum Corona-Diskurs skizziert, die für die Beispielanalyse relevant sind. Exemplarisch werden in der vorliegenden Untersuchung redaktionelle Beiträge und Fake News zum Corona-Diskurs qualitative analysiert und miteinander verglichen. Damit wird der Versuch unternommen Fake News als Textsorte einzuordnen und die sprachlichen Strategien herauszuarbeiten, die für dieses Phänomen typisch sind. Die zuvor erwähnten linguistischen Zugänge liefern dabei wertvolle Ergebnisse, die als Ausgangsbasis für diese Untersuchung dienen. Die Analyse zeigt, dass sich Fake News zwar an journalistischen Nachrichtenformaten orientieren, jedoch sprachliche Muster aufweisen, die sie von diesen unterscheiden. Unter anderem benutzen sie sprachliche Strategien wie Polarisierung, starke Wertungen und direkte Leseradressierung, die für redaktionelle Berichterstattung unüblich ist. Diese Strategien sind nicht nur für Fake News typisch, sondern weisen eine gewisse Parallele zu sprachlichen Mustern in Verschwörungstheorien und Propaganda auf. [less ▲]

Detailed reference viewed: 58 (0 UL)
Full Text
Peer Reviewed
See detailA Machine Learning Approach for Automated Filling of Categorical Fields in Data Entry Forms
Belgacem, Hichem UL; Li, Xiaochen; Bianculli, Domenico UL et al

in ACM Transactions on Software Engineering and Methodology (in press)

Users frequently interact with software systems through data entry forms. However, form filling is time-consuming and error-prone. Although several techniques have been proposed to auto-complete or pre ... [more ▼]

Users frequently interact with software systems through data entry forms. However, form filling is time-consuming and error-prone. Although several techniques have been proposed to auto-complete or pre-fill fields in the forms, they provide limited support to help users fill categorical fields, i.e., fields that require users to choose the right value among a large set of options. In this paper, we propose LAFF, a learning-based automated approach for filling categorical fields in data entry forms. LAFF first builds Bayesian Network models by learning field dependencies from a set of historical input instances, representing the values of the fields that have been filled in the past. To improve its learning ability, LAFF uses local modeling to effectively mine the local dependencies of fields in a cluster of input instances. During the form filling phase, LAFF uses such models to predict possible values of a target field, based on the values in the already-filled fields of the form and their dependencies; the predicted values (endorsed based on field dependencies and prediction confidence) are then provided to the end-user as a list of suggestions. We evaluated LAFF by assessing its effectiveness and efficiency in form filling on two datasets, one of them proprietary from the banking domain. Experimental results show that LAFF is able to provide accurate suggestions with a Mean Reciprocal Rank value above 0.73. Furthermore, LAFF is efficient, requiring at most 317 ms per suggestion. [less ▲]

Detailed reference viewed: 65 (12 UL)
Full Text
Peer Reviewed
See detailEmpirical risk assessment of maintenance costs under full-service contracts
Deprez, Laurens UL; Antonio, Katrien; Boute, Robert

in European Journal of Operational Research (in press)

We provide a data-driven framework to conduct a risk assessment, including data pre-processing, exploration, and statistical modeling, on a portfolio of full-service maintenance contracts. These contracts ... [more ▼]

We provide a data-driven framework to conduct a risk assessment, including data pre-processing, exploration, and statistical modeling, on a portfolio of full-service maintenance contracts. These contracts cover all maintenance-related costs for a fixed, upfront fee during a predetermined horizon. Charging each contract a price proportional to its risk prevents adverse selection by incentivizing low risk (i.e., maintenance-light) profiles to not renege on their agreements. We borrow techniques from non-life insurance pricing and tailor them to the setting of maintenance contracts to assess the risk and estimate the expected maintenance costs under a full-service contract. We apply the framework on a portfolio of about 5 000 full-service contracts of industrial equipment and show how a data-driven analysis based on contract and machine characteristics, or risk factors, supports a differentiated, risk-based break-even tariff plan. We employ generalized additive models (GAMs) to predict the risk factors’ impact on the frequency (number of) and severity (cost) of maintenance interventions. GAMs are interpretable yet flexible statistical models that capture the effect of both continuous and categorical risk factors. Our predictive models quantify the impact of the contract and machine type, service history, and machine running hours on the contract cost. We additionally utilize the predictive cost distributions of our models to augment the break-even price with the appropriate risk margins to further protect against the inherently stochastic nature of the maintenance costs. The framework shows how maintenance intervention data can set up a differentiated tariff plan. [less ▲]

Detailed reference viewed: 26 (0 UL)
Full Text
Peer Reviewed
See detailIntercorrelated random fields with bounds and the Bayesian identification of their parameters: Application to linear elastic struts and fibers
Rappel, Hussein; Girolami, Mark; Beex, Lars UL

in International Journal for Numerical Methods in Engineering (in press)

Many materials and structures consist of numerous slender struts or fibers. Due to the manufacturing processes of different types of struts and the growth processes of natural fibers, their mechanical ... [more ▼]

Many materials and structures consist of numerous slender struts or fibers. Due to the manufacturing processes of different types of struts and the growth processes of natural fibers, their mechanical response frequently fluctuates from strut to strut, as well as locally within each strut. In associated mechanical models each strut is often represented by a string of beam elements, since the use of conventional three-dimensional finite elements renders the simulations computationally inefficient. The parameter input fields of each string of beam elements are ideally such that the local fluctuations and fluctuations between individual strings of beam elements are accurately captured. The goal of this study is to capture these fluctuations in several intercorrelated bounded random fields. Two formulations to describe the intercorrelations between each random field, as well as the case without any intercorrelation, are investigated. As only a few sets of input fields are available (due to time constraints of the supposed experimental techniques), the identification of the random fields’ parameters involves substantial uncertainties. A probabilistic identification approach based on Bayes’ theorem is employed to treat the involved uncertainties. [less ▲]

Detailed reference viewed: 39 (0 UL)
Full Text
Peer Reviewed
See detailCOVID-19 and the global venture capital landscape
Bellavitis, Cristiano; Fisch, Christian UL; McNaughton, Rod

in Small Business Economics (in press)

We assess the effect of the COVID-19 pandemic on venture capital (VC) investments, documenting a significant decline in investments using a dataset of 39,527 funding rounds occurring before and during the ... [more ▼]

We assess the effect of the COVID-19 pandemic on venture capital (VC) investments, documenting a significant decline in investments using a dataset of 39,527 funding rounds occurring before and during the pandemic in 130 countries. In line with our theoretical considerations, we show that this decline is more pronounced for investments characterized by higher uncertainty, namely investments in seed-stage ventures, industries affected more heavily by the COVID-19 crisis, international investments, and non-syndicated investments. Investor prominence partially moderates these effects. [less ▲]

Detailed reference viewed: 21 (2 UL)
Full Text
Peer Reviewed
See detailAutomated Question Answering for Improved Understanding of Compliance Requirements: A Multi-Document Study
Abualhaija, Sallam UL; Arora, Chetan; Sleimi, Amin et al

in In Proceedings of the 30th IEEE International Requirements Engineering Conference (RE'22), Melbourne, Australia 15-19 August 2022 (in press)

Software systems are increasingly subject to regulatory compliance. Extracting compliance requirements from regulations is challenging. Ideally, locating compliance-related information in a regulation ... [more ▼]

Software systems are increasingly subject to regulatory compliance. Extracting compliance requirements from regulations is challenging. Ideally, locating compliance-related information in a regulation requires a joint effort from requirements engineers and legal experts, whose availability is limited. However, regulations are typically long documents spanning hundreds of pages, containing legal jargon, applying complicated natural language structures, and including cross-references, thus making their analysis effort-intensive. In this paper, we propose an automated question-answering (QA) approach that assists requirements engineers in finding the legal text passages relevant to compliance requirements. Our approach utilizes large-scale language models fine-tuned for QA, including BERT and three variants. We evaluate our approach on 107 question-answer pairs, manually curated by subject-matter experts, for four different European regulatory documents. Among these documents is the general data protection regulation (GDPR) – a major source for privacy-related requirements. Our empirical results show that, in ~94% of the cases, our approach finds the text passage containing the answer to a given question among the top five passages that our approach marks as most relevant. Further, our approach successfully demarcates, in the selected passage, the right answer with an average accuracy of ~ 91%. [less ▲]

Detailed reference viewed: 138 (5 UL)
Full Text
Peer Reviewed
See detailOptimal Priority Assignment for Real-Time Systems: A Coevolution-Based Approach
Lee, Jaekwon UL; Shin, Seung Yeob UL; Nejati, Shiva et al

in Empirical Software Engineering (in press)

In real-time systems, priorities assigned to real-time tasks determine the order of task executions, by relying on an underlying task scheduling policy. Assigning optimal priority values to tasks is ... [more ▼]

In real-time systems, priorities assigned to real-time tasks determine the order of task executions, by relying on an underlying task scheduling policy. Assigning optimal priority values to tasks is critical to allow the tasks to complete their executions while maximizing safety margins from their specified deadlines. This enables real-time systems to tolerate unexpected overheads in task executions and still meet their deadlines. In practice, priority assignments result from an interactive process between the development and testing teams. In this article, we propose an automated method that aims to identify the best possible priority assignments in real-time systems, accounting for multiple objectives regarding safety margins and engineering constraints. Our approach is based on a multi-objective, competitive coevolutionary algorithm mimicking the interactive priority assignment process between the development and testing teams. We evaluate our approach by applying it to six industrial systems from different domains and several synthetic systems. The results indicate that our approach significantly outperforms both our baselines, i.e., random search and sequential search, and solutions defined by practitioners. Our approach scales to complex industrial systems as an offline analysis method that attempts to find near-optimal solutions within acceptable time, i.e., less than 16 hours. [less ▲]

Detailed reference viewed: 55 (18 UL)
Full Text
Peer Reviewed
See detailEstimating Probabilistic Safe WCET Ranges of Real-Time Systems at Design Stages
Lee, Jaekwon UL; Shin, Seung Yeob UL; Nejati, Shiva et al

in ACM Transactions on Software Engineering and Methodology (in press)

Estimating worst-case execution times (WCET) is an important activity at early design stages of real-time systems. Based on WCET estimates, engineers make design and implementation decisions to ensure ... [more ▼]

Estimating worst-case execution times (WCET) is an important activity at early design stages of real-time systems. Based on WCET estimates, engineers make design and implementation decisions to ensure that task execution always complete before their specified deadlines. However, in practice, engineers often cannot provide precise point WCET estimates and prefer to provide plausible WCET ranges. Given a set of real-time tasks with such ranges, we provide an automated technique to determine for what WCET values the system is likely to meet its deadlines, and hence operate safely with a probabilistic guarantee. Our approach combines a search algorithm for generating worst-case scheduling scenarios with polynomial logistic regression for inferring probabilistic safe WCET ranges. We evaluated our approach by applying it to three industrial systems from different domains and several synthetic systems. Our approach efficiently and accurately estimates probabilistic safe WCET ranges within which deadlines are likely to be satisfied with a high degree of confidence. [less ▲]

Detailed reference viewed: 46 (5 UL)
Peer Reviewed
See detailLearning and Teaching Geropsychology
Boll, Thomas UL

in Zumbach, Joerg; Bernstein, Douglas; Narciss, Susanne (Eds.) et al International Handbook of Psychology Learning and Teaching (in press)

This chapter presents a broad overview about the essential aspects of learning and teaching geropsychology in tertiary education. After an introduction to the scope of geropsychology and the need for ... [more ▼]

This chapter presents a broad overview about the essential aspects of learning and teaching geropsychology in tertiary education. After an introduction to the scope of geropsychology and the need for geropsychology education, the objectives and basic principles of geropsychology curricula are discussed. Their overall goal is to qualify students and/or professionals to understand and solve psychological aspects of problems of older people in the contexts of practical application, research, and teaching. Specific teaching and learning objectives in geropsychology are described in terms of underlying content dimensions (e.g., areas of acting, functioning, and development of older people; basic components of practical geropsychological acting; target groups and settings) and levels of competency to be acquired (e.g., uni-, multi-structural, relational, and extended abstract level). This is followed by an overview of the core topics of geropsychology. These refer to theoretical (e.g., basic concepts of age, aging and the elderly; action competence of older people; challenges in later life; resources for adaptation; problems of people providing services to older adults), normative (ethical and legal) and methodological foundations (research, assessment, evaluation and intervention methods). This is followed by sections on linking main learning objectives and core topics of geropsychology to courses within study programs (psychological or non-psychological) and about the relations between teaching, learning, and assessment in geropsychology. The chapter concludes with information about resources for these issues including relevant URL links, tips for teaching and annotated references to further reading. [less ▲]

Detailed reference viewed: 47 (7 UL)
Full Text
Peer Reviewed
See detailReplication studies in top management journals: an empirical investigation of prevalence, types, outcomes, and impact
Block, Jörn; Fisch, Christian UL; Kanwal, Narmeen et al

in Management Review Quarterly (in press)

Replication studies are important for the empirical research process. Yet, while there is an increased awareness of the need for replication in management research, it appears that such studies are rarely ... [more ▼]

Replication studies are important for the empirical research process. Yet, while there is an increased awareness of the need for replication in management research, it appears that such studies are rarely published in leading management journals. Importantly, we lack a comprehensive overview of replication studies in the top management journals that spans all sub-disciplines. Our systematic review closes this gap and provides an overview of the prevalence, types, outcomes, and impact of replication studies in management journals. We find that differences in the prevalence of replications between sub-disciplines exist and that most replications are wide replications. With regard to the replication outcome, our review shows that the share of non-confirming replications is low. Moreover, such replications are cited less often than confirming replications pointing towards a confirmation bias in management research. We discuss the implications of our results for authors, reviewers, and editors of management journals. [less ▲]

Detailed reference viewed: 23 (1 UL)
Full Text
Peer Reviewed
See detailNo Ensemble Averaging Below the Black Hole Threshold
Schlenker, Jean-Marc UL; Witten, Edward

in Journal of High Energy Physics (in press)

In the AdS/CFT correspondence, amplitudes associated to connected bulk manifolds with disconnected boundaries have presented a longstanding mystery. A possible interpretation is that they reflect the ... [more ▼]

In the AdS/CFT correspondence, amplitudes associated to connected bulk manifolds with disconnected boundaries have presented a longstanding mystery. A possible interpretation is that they reflect the effects of averaging over an ensemble of boundary theories. But in examples in dimension D≥3, an appropriate ensemble of boundary theories does not exist. Here we sharpen the puzzle by identifying a class of "sub-threshold" observables that we claim do not show effects of ensemble averaging. These are amplitudes that do not involve black hole states. To support our claim, we explore the example of D=3, and show that connected solutions of Einstein's equations with disconnected boundary never contribute to sub-threshold observables. To demonstrate this requires some novel results about the renormalized volume of a hyperbolic three-manifold, which we prove using modern methods in hyperbolic geometry. Why then do any observables show apparent ensemble averaging? We propose that this reflects the chaotic nature of black hole physics and the fact that the Hilbert space describing a black hole does not have a large N limit. [less ▲]

Detailed reference viewed: 5 (0 UL)
Full Text
Peer Reviewed
See detailIntelligent Blockchain-based Edge Computing via Deep Reinforcement Learning: Solutions and Challenges
Nguyen, Dinh C; Nguyen, van Dinh UL; Ding, Ming et al

in IEEE Network (in press)

The convergence of mobile edge computing (MEC) and blockchain is transforming the current computing services in wireless Internet-of-Things networks, by enabling task offloading with security enhancement ... [more ▼]

The convergence of mobile edge computing (MEC) and blockchain is transforming the current computing services in wireless Internet-of-Things networks, by enabling task offloading with security enhancement based on blockchain mining. Yet the existing approaches for these enabling technologies are isolated, providing only tailored solutions for specific services and scenarios. To fill this gap, we propose a novel cooperative task offloading and blockchain mining (TOBM) scheme for a blockchain-based MEC system, where each edge device not only handles computation tasks but also deals with block mining for improving system utility. To address the latency issues caused by the blockchain operation in MEC, we develop a new Proof-of-Reputation consensus mechanism based on a lightweight block verification strategy. To accommodate the highly dynamic environment and high-dimensional system state space, we apply a novel distributed deep reinforcement learning-based approach by using a multi-agent deep deterministic policy gradient algorithm. Experimental results demonstrate the superior performance of the proposed TOBM scheme in terms of enhanced system reward, improved offloading utility with lower blockchain mining latency, and better system utility, compared to the existing cooperative and non-cooperative schemes. The paper concludes with key technical challenges and possible directions for future blockchain-based MEC research. [less ▲]

Detailed reference viewed: 11 (2 UL)
Full Text
Peer Reviewed
See detailCorporate Managers, Price Noise and the Investment Factor
Lehnert, Thorsten UL

in Financial Innovation (2022), 8(61),

Detailed reference viewed: 42 (7 UL)
Full Text
Peer Reviewed
See detailBetting Against Noisy Beta
Lehnert, Thorsten UL

in Journal of Finance and Data Science (2022), 8(-), 55-68

Detailed reference viewed: 44 (4 UL)