87,085 research outputs found
Distributed reasoning for the autonomous coordination of smart object networks
Full text linkL'evoluzione dell'Internet of Things (IoT) verso l'Internet of Everything (IoE) riflette il progresso dei dispositivi di rete e delle tecnologie di calcolo, comprendendo non solo oggetti, ma anche ambienti, persone, processi e dati. Questo sviluppo consente una raccolta e un'analisi dei dati su larga scala, con il potenziale di trasformare le interazioni tra molteplici attività umane e il mondo fisico. Sebbene questa transizione migliori l'efficienza operativa e il processo decisionale basato sui dati, la sua piena realizzazione richiede il superamento di problematiche relative alla larghezza di banda di rete, al consumo energetico, alla sicurezza dei dati e alla privacy. Soprattutto, nell'IoE, l'interoperabilità e la gestione intelligente delle informazioni diventano fondamentali per supportare processi autonomi flessibili e architetture orientate ai servizi sofisticate, adatte a interazioni estese tra macchine e tra esseri umani e macchine.
Una strategia chiave per affrontare queste sfide è l’edge computing, che avvicina le attività computazionali alle sorgenti di dati. Questa trasformazione è essenziale per gestire i grandi volumi di dati e la rapidità con cui questi sono generati nell'IoE, mitigando al contempo i problemi di latenza e larghezza di banda associati ai sistemi di elaborazione centralizzata. Un primo esempio di framework intelligente che sfrutta l’edge computing è il Semantic Web of Things (SWoT). In questo contesto, descrizioni basate sull’utilizzo di ontologie di dispositivi, oggetti ed eventi vengono gestite localmente da agenti intelligenti pervasivi attraverso ragionamenti automatizzati, consentendo operazioni autonome orientate a obiettivi specifici.
L'avanzamento del SWoT verso un Semantic Web of Everything (SWoE) richiede un'integrazione più profonda delle tecnologie semantiche nelle interazioni di calcolo pervasivo. Questa visione implica una pervasività di strumenti di rappresentazione della conoscenza e capacità di ragionamento automatizzato, anche su dispositivi con limitate capacità di elaborazione, memoria ed energia. Meccanismi di inferenza locale sui dispositivi sono essenziali nello SWoE, considerando l'elevata volatilità e la limitata accessibilità a dispositivi più performanti.
L'implementazione di architetture SWoE presenta difficoltà significative dal punto di vista scientifico e tecnologico. Gli attuali motori di ragionamento per il Semantic Web e i Knowledge Base Management Systems (KBMS) sono principalmente progettati per ambienti di calcolo ad elevate prestazioni, come server e cluster di workstation, rendendoli inadatti a dispositivi su scala nanometrica. I motori di ragionamento che potrebbero funzionare su dispositivi più piccoli spesso mancano di procedure di inferenza essenziali, limitandone l'utilizzo. Per questo motivo, la creazione di piattaforme SWoE richiede una rivalutazione delle metodologie di valutazione e benchmarking per includere i vincoli unici di questo nuovo paradigma.
Questa dissertazione presenta diversi contributi innovativi nel campo del ragionamento distribuito in scenari SWoE, concentrandosi sull'applicazione della rappresentazione della conoscenza e del ragionamento automatizzato al coordinamento di reti di agenti intelligenti incorporati in dispositivi dalle risorse limitate. A tal fine, questo lavoro analizza architetture e strategie di ottimizzazione per vari componenti fondamentali, come: Cowl, una libreria per la rappresentazione della conoscenza leggera e versatile progettata per dispositivi dalle risorse limitate, che supera le restrizioni dei KBMS attuali nei contesti embedded e IoT; Tiny-ME, un innovativo motore di ragionamento e matchmaking multi-piattaforma progettato per lo SWoE, che offre capacità di ragionamento efficienti adatte a dispositivi cloud, desktop, mobili ed edge; evOWLuator, un framework multipiattaforma per il benchmarking di motori di ragionamento del Semantic Web, con enfasi sulla stima del consumo energetico e sul supporto inferenziale su dispositivi remoti; un framework di Cloud-Edge Intelligence (CEI) per sistemi multi-agente e applicazioni basate su sensori, che sfrutta il calcolo serverless per la gestione dei dati e i task di machine learning.
Grande enfasi è posta sulla valutazione delle tecnologie sviluppate attraverso campagne sperimentali estese, che forniscono approfondimenti su prestazioni, efficienza e applicabilità in contesti SWoE. Inoltre, vengono dimostrate applicazioni pratiche attraverso casi di studio in diversi contesti. Il primo scenario presenta un framework per l'adattamento della Quality of Experience (QoE) nello streaming multimediale Web, utilizzando la versione WebAssembly di Tiny-ME come motore di ragionamento. Il secondo evidenzia un sistema di ricerca di eventi locali incentrato sulla privacy, mostrando un caso d'uso di ragionamento client-side per il recupero e la personalizzazione dei dati in applicazioni Web. Il terzo esempio esplora come Tiny-ME è in grado di gestire risorse annotate semanticamente in reti peer-to-peer, migliorando negoziazioni e l’explanation dei risultati di ricerca. Infine, un esempio di smart city mostra come Cowl può essere integrato in sensori su scala nanometrica per lo scambio di dati arricchiti semanticamente, migliorando la mobilità urbana. Gli esperimenti e le applicazioni menzionati evidenziano la flessibilità e la vasta applicabilità dei metodi e delle tecnologie presentati, sottolineando il potenziale esteso dello SWoE.The evolution of the Internet of Things (IoT) into the Internet of Everything (IoE) reflects the evolution of connected devices and computing technologies, encompassing not only things, but also environments, people, processes, and data. It enables large-scale data collection and analytics, with the potential to transform the interactions between many kinds of human activities and the physical world. Although this transition improves operational efficiency and data-driven decision-making, its full realization requires overcoming issues concerning network bandwidth, energy consumption, data security, and privacy. Most importantly, in the IoE interoperability and smart information management become essential for supporting flexible autonomous processing and sophisticated, flexible service-oriented architectures for extensive machine-to-machine and human-machine interactions.
A key strategy for addressing these challenges is edge computing, which brings computational tasks closer to data sources. This transformation is essential for managing the large volumes and rapid pace of data generated in the IoE, while also mitigating latency and bandwidth issues associated with centralized processing systems. An early example of a smart framework that leverages edge computing is the Semantic Web of Things (SWoT). Here, ontology-based descriptions of devices, objects, and events are dealt with locally by pervasive intelligent agents through automated reasoning, enabling autonomous operations towards specific objectives.
The advancement of SWoT towards a Semantic Web of Everything (SWoE) requires deeper embedding of semantic technologies in pervasive computing interactions. This vision requires pervasive knowledge representation and automated reasoning abilities, even on devices with stringent processing, memory, and energy limitations. Local inference mechanisms on devices are essential in the SWoE, considering the high volatility and restricted accessibility of more powerful devices.
The deployment of SWoE architectures discloses considerable difficulties from a scientific and technological standpoint. Current Semantic Web reasoners and Knowledge Base Management Systems (KBMS) are primarily tailored for high-performance computing environments such as servers and workstation clusters, making them unsuitable for nano-scale devices. Reasoning engines that might work on smaller devices frequently lack essential inference support, thus limiting their practicality. For this reason, creating SWoE platforms requires a reassessment of evaluation and benchmarking methodologies to consider the unique constraints of this new paradigm.
This dissertation presents several innovative contributions to the field of distributed reasoning in SWoE scenarios, focusing on applying knowledge representation and automated inferences to the coordination of networks of smart agents embodied into resource-constrained devices. To this aim, this work covers system architectures and optimization strategies for various essential components frameworks, such as: Cowl, a lightweight and versatile knowledge representation library designed for devices with limited resources, overcoming the restrictions of current KBMS in embedded and IoT contexts; Tiny-ME, an innovative multi-platform reasoner and matchmaking engine tailored for the SWoE, providing efficient reasoning capabilities appropriate for cloud, desktop, mobile, and edge devices; evOWLuator, a cross-platform evaluation framework that is mindful of energy consumption for Semantic Web reasoners, emphasizing power usage estimation and supporting inferences on remote devices; a Cloud-Edge Intelligence (CEI) framework for multi-agent systems and sensor-based application, exploiting serverless computing for data management and machine learning tasks.
Great emphasis is placed on the assessment of the developed technologies through extensive experimental campaigns, which provide insights into performance, efficiency, and applicability in SWoE settings. In addition, practical applications are demonstrated through case studies in various contexts. The first scenario demonstrates a framework for adapting Quality of Experience (QoE) in Web multimedia streaming, using the WebAssembly port of Tiny-ME as reasoning engine. The second highlights a privacy-focused local event finder, showing a client-side Web reasoning use case in data retrieval and personalization for Web applications. The third case study explores how Tiny-ME manages semantically annotated resources in peer-to-peer networks, improving negotiation and discovery explanations. Finally, a smart city example shows how Cowl can be integrated in nano-scale sensors to exchange semantically enriched data, enhancing urban mobility. Together, the mentioned experiments and applications underscore the flexibility and wide-ranging usability of the presented methods and technologies, highlighting the extensive potential of the SWoE
The Folio: The Magazine of Forman Christian College
No full textEditorial. pp. 3-4; Velte, Mowbray-Dr Bashir Ahmad. pp. 5; Saeed Karim Fazli-Mr Boyce: an Appreciation. pp. 6-8; Siraj-ud-Din-Speech-Valedictory Address: Delivered at the College Assembly Hall on 28th March. pp. 9-13; Basil P. Das-Article-Muslim Architecture. pp. 14-16; Riaz Hussain-Article-The Contribution of European Writers. pp. 17-21; Robbins, S. W.-Article-A Modern Approach to English Poetry. pp. 22-33; Mackenize, Donald G.-Poetry-I am a Nation. pp. 33; Saeed Ahmad-Article-Lyric Poetry. pp. 34-37; Aijaz ul Haque-Article-The Novels of Thomas Hardy. pp. 38-40; Eshtiaq A. Siddiqui-Poetry-God or Gods?. pp. 40; Najm Hussain Syed-Article-Humanity in the Plays of Galsworthy. pp. 41-42; Wisal Khan-Article-Sir Winston Churchill. pp. 43-45; Zia ur Rahman-On Going Hunting. pp. 46-48; Saeed Karin Fazli-The Leisure Way. pp. 49; Aiyaz ul Haq-Story-The Skirt Girl. pp. 51-54; Aftab A. Jan-Story-Men from Venus. pp. 55-57; Mohd Zafar Khattak-Story-Shahnaz. pp. 58-60; Malik, M. Naseem A.-Story-The Coat. pp. 61-63; Saeed Akhtar-Story-Love is a Many Splendoured Thing. pp. 64-66; The Societies Report. pp. 67-70; Saeed Ahmad-Poetry-The Blue-Bells Toll for Thee. pp. 72; Velte, F. M.-F. C. College Sports, 1956-57. pp. 73-76; Folio [Urdu]. 58 p.Editorial Board 1957. before Editorial page; Dr F. Mowbray Velte. after page 18; Mr Stanley E. Brush, Izharuddin Ahmed (President, College Union), Iftikhar Gilani (President, Secondary Union). after page 34; Arthur Mervyn (Valedictory Address, delivered at the F.C. College Hall on 28th March), Hamayun Khan Afredi (Captain of College Football Team), Ijaz Akhar (Captain of Degree Basket Ball Team). after page 50; The F.c.c. Secondary Board Basketball Team. after page 6
Efficient Frequency and Time-Domain Simulations of Delayed PEEC Models With Proper Orthogonal Decomposition Techniques
Full text linkThe Partial Element Equivalent Circuit (PEEC) method has gained significant recognition as an electromagnetic computational technique known for its ability to represent electromagnetic phenomena using equivalent circuits. This feature makes it particularly valuable for addressing mixed EM-circuit problems. However, PEEC models often exhibit large dimensions, necessitating modeling techniques that can effectively reduce their size while preserving accuracy. Model order reduction (MOR) serves as a highly effective approach to accomplish this objective. This paper presents two MOR techniques based on proper orthogonal decomposition (POD) for PEEC models described by neutral delayed differential equations (NDDEs). The unique characteristics of NDDEs demand specialized MOR approaches, as their formulation is inherently more complex compared to standard quasi-static PEEC models described by non-delayed differential equations. In addition to a traditional one-shot singular value decomposition (SVD), this paper also presents an incrementally computed SVD to evaluate the orthogonal matrix needed to generate the reduced order matrices. The accuracy and efficiency of the proposed PEEC-MOR methods are demonstrated through multiple relevant numerical results in both the frequency-domain and time-domain
Derivatives-Enhanced Proper Orthogonal Decomposition for PEEC Models With Delays
No full textThis letter proposes a novel model order reduction (MOR) approach leveraging frequency-domain proper orthogonal decomposition (POD) for partial element equivalent circuit (PEEC) models characterized by neutral delayed differential equations (NDDEs). Our technique incorporates frequency-domain derivatives snapshots alongside frequency-domain response snapshots, thereby enhancing the accuracy of the reduced-order model while minimizing the computational overhead compared with solely utilizing frequency-domain response snapshots. A numerical example is provided to demonstrate the effectiveness and efficiency of the proposed method in both the frequency domain and the time domain
A microencapsulated blend of botanicals and organic acids can improve growth performance and gut health of broilers
No full textRecommended from our members
Toward a Robust and Universal Crowd Labeling Framework
Full text linkThe advent of fast and economical computers with large electronic storage has led to a large volume of data, most of which is unlabeled. While computers provide expeditious, accurate and low-cost computation, they still lag behind in many tasks that require human intelligence such as labeling medical images, videos or text. Consequently, current research focuses on a combination of computer accuracy and human intelligence to complete labeling task. In most cases labeling needs to be done by domain experts, however, because of the variability in expertise, experience, and intelligence of human beings, experts can be scarce.
As an alternative to using domain experts, help is sought from non-experts, also known as Crowd, to complete tasks that cannot be readily automated. Since crowd labelers are non-expert, multiple labels per instance are acquired for quality purposes. The final label is obtained by com- bining these multiple labels. It is very common that the ground truth, instance difficulty, and the labeler ability are unknown entities. Therefore, the aggregation task becomes a “chicken and egg” problem to start with.
Despite the fact that much research using machine learning and statistical techniques has been conducted in this area (e.g., [Dekel and Shamir, 2009; Hovy et al., 2013a; Liu et al., 2012; Donmez and Carbonell, 2008]), many questions remain unresolved, these include: (a) What are the best ways to evaluate labelers? (b) It is common to use expert-labeled instances (ground truth) to evaluate la- beler ability (e.g., [Le et al., 2010; Khattak and Salleb-Aouissi, 2011; Khattak and Salleb-Aouissi, 2012; Khattak and Salleb-Aouissi, 2013]). The question is, what should be the cardinality of the set of expert-labeled instances to have an accurate evaluation? (c) Which factors other than labeler expertise (e.g., difficulty of instance, prevalence of class, bias of a labeler toward a particular class) can affect the labeling accuracy? (d) Is there any optimal way to combine multiple labels to get the
best labeling accuracy? (e) Should the labels provided by oppositional/malicious labelers be dis- carded and blocked? Or is there a way to use the “information” provided by oppositional/malicious labelers? (f) How can labelers and instances be evaluated if the ground truth is not known with certitude?
In this thesis, we investigate these questions. We present methods that rely on few expert-labeled instances (usually 0.1% -10% of the dataset) to evaluate various parameters using a frequentist and a Bayesian approach. The estimated parameters are then used for label aggregation to produce one final label per instance.
In the first part of this thesis, we propose a method called Expert Label Injected Crowd Esti- mation (ELICE) and extend it to different versions and variants. ELICE is based on a frequentist approach for estimating the underlying parameters. The first version of ELICE estimates the pa- rameters i.e., labeler expertise and data instance difficulty, using the accuracy of crowd labelers on expert-labeled instances [Khattak and Salleb-Aouissi, 2011; Khattak and Salleb-Aouissi, 2012]. The multiple labels for each instance are combined using weighted majority voting. These weights are the scores of labeler reliability on any given instance, which are obtained by inputting the pa- rameters in the logistic function.
In the second version of ELICE [Khattak and Salleb-Aouissi, 2013], we introduce entropy as a way to estimate the uncertainty of labeling. This provides an advantage of differentiating between good, random and oppositional/malicious labelers. The aggregation of labels for ELICE version 2 flips the label (for binary classification) provided by the oppositional/malicious labeler thus utilizing the information that is generally discarded by other labeling methodologies.
Both versions of ELICE have a cluster-based variant in which rather than making a random choice of instances from the whole dataset, clusters of data are first formed using any clustering approach e.g., K-means. Then an equal number of instances from each cluster are chosen randomly to get expert-labels. This is done to ensure equal representation of each class in the test dataset.
Besides taking advantage of expert-labeled instances, the third version of ELICE [Khattak and Salleb-Aouissi, 2016], incorporates pairwise/circular comparison of labelers to labelers and in- stances to instances. The idea here is to improve accuracy by using the crowd labels, which unlike expert-labels, are available for the whole dataset and may provide a more comprehensive view of the labeler ability and instance difficulty. This is especially helpful for the case when the domain
experts do not agree on one label and ground truth is not known for certain. Therefore, incorporating more information beyond expert labels can provide better results.
We test the performance of ELICE on simulated labels as well as real labels obtained from Amazon Mechanical Turk. Results show that ELICE is effective as compared to state-of-the-art methods. All versions and variants of ELICE are capable of delaying phase transition. The main contribution of ELICE is that it makes the use of all possible information available from crowd and experts. Next, we also present a theoretical framework to estimate the number of expert-labeled instances needed to achieve certain labeling accuracy. Experiments are presented to demonstrate the utility of the theoretical bound.
In the second part of this thesis, we present Crowd Labeling Using Bayesian Statistics (CLUBS) [Khattak and Salleb-Aouissi, 2015; Khattak et al., 2016b; Khattak et al., 2016a], a new approach for crowd labeling to estimate labeler and instance parameters along with label aggregation. Our approach is inspired by Item Response Theory (IRT). We introduce new parameters and refine the existing IRT parameters to fit the crowd labeling scenario. The main challenge is that unlike IRT, in the crowd labeling case, the ground truth is not known and has to be estimated based on the parameters. To overcome this challenge, we acquire expert-labels for a small fraction of instances in the dataset. Our model estimates the parameters based on the expert-labeled instances. The estimated parameters are used for weighted aggregation of crowd labels for the rest of the dataset. Experiments conducted on synthetic data and real datasets with heterogeneous quality crowd-labels show that our methods perform better than many state-of-the-art crowd labeling methods.
We also conduct significance tests between our methods and other state-of-the-art methods to check the significance of the accuracy of these methods. The results show the superiority of our method in most cases. Moreover, we present experiments to demonstrate the impact of the accuracy of final aggregated labels when used as training data. The results essentially emphasize the need for high accuracy of the aggregated labels.
In the last part of the thesis, we present past and contemporary research related to crowd la- beling. We conclude with future of crowd labeling and further research directions. To summarize, in this thesis, we have investigated different methods for estimating crowd labeling parameters and using them for label aggregation. We hope that our contribution will be useful to the crowd labeling community
Proper Orthogonal Decomposition-Based Model Order Reduction of Delayed PEEC Models
No full textThe Partial Element Equivalent Circuit (PEEC) method is an electromagnetic computational method that has attracted a lot of attention for its capability to represent electromagnetic phenomena by equivalent circuits. This allows combining mixed EM-circuit problems in a straightforward manner, which proves to be very useful for mixed EM-circuit problems. However, the PEEC models can result of large size and therefore it becomes important to have modeling techniques that can compress the size of these models while retaining accuracy. Model order reduction (MOR) is a very effective way to achieve this goal. In this paper, we present a proper orthogonal decomposition (POD) based MOR for PEEC models that are described by neutral delayed differential equations (NDDEs). NDDEs require dedicated MOR schemes since the form of those equations is definitely more complex that standard quasi-static PEEC models described by standard (non-delayed) differential equations. The proposed PEEC-MOR method is validated by pertinent numerical results
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
The Possible Role of Prescribing Medications, Including Central Nervous System Drugs, in Contributing to Male-Factor Infertility (MFI): Assessment of the Food and Drug Administration (FDA) Pharmacovigilance Database
Full text linkBackground: A wide range of medications may have a possible role in the development of male-factor infertility (MFI), including various antineoplastic agents, testosterone/anabolic steroids, immunosuppressive drugs/immunomodulators, glucocorticosteroids, non-steroidal anti-inflammatory drugs, opiates, antiandrogenic drugs/5-alpha-reductase inhibitors, various antibiotics, antidepressants, antipsychotics, antiepileptic agents and others. We aimed at investigating this issue from a pharmacovigilance-based perspective. Methods: The Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS) database was queried to identify the drugs associated the most with MFI individual reports. Only those drugs being associated with more than 10 MFI reports were considered for the disproportionality analysis. Proportional Reporting Ratios (PRRs) and their confidence intervals were computed for all the drugs identified in this way in January 2023. Secondary, ‘unmasking’, dataset analyses were carried out as well. Results: Out of the whole database, 955 MFI reports were identified, 408 (42.7%) of which were associated with 20 medications, which had more than 10 reports each. Within this group, finasteride, testosterone, valproate, diethylstilbestrol, mechloretamine, verapamil, lovastatin and nifedipine showed significant levels of actual disproportionate reporting. Out of these, and before unmasking, the highest PRR values were identified for finasteride, diethylstilbestrol and mechloretamine, respectively, with values of 16.0 (12.7–20.3), 14.3 (9.1–22.4) and 58.7 (36.3–95.9). Conclusions: A variety of several medications, a number of which were already supposed to be potentially linked with MFI based on the existing evidence, were associated with significant PRR levels for MFI in this analysis. A number of agents which were previously hypothesized to be associated with MFI were not represented in this analysis, suggesting that drug-induced MFI is likely under-reported to regulatory agencies. Reproductive medicine specialists should put more effort into the detection and reporting of these adverse drug reactions
- …
