• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 71
  • 13
  • 7
  • 5
  • 4
  • 3
  • 3
  • 3
  • 2
  • 1
  • Tagged with
  • 139
  • 139
  • 32
  • 31
  • 30
  • 27
  • 22
  • 22
  • 20
  • 20
  • 19
  • 19
  • 18
  • 18
  • 16
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
91

Inference in Generalized Linear Models with Applications

Byrne, Evan 29 August 2019 (has links)
No description available.
92

A Bridge between Graph Neural Networks and Transformers: Positional Encodings as Node Embeddings

Manu, Bright Kwaku 01 December 2023 (has links) (PDF)
Graph Neural Networks and Transformers are very powerful frameworks for learning machine learning tasks. While they were evolved separately in diverse fields, current research has revealed some similarities and links between them. This work focuses on bridging the gap between GNNs and Transformers by offering a uniform framework that highlights their similarities and distinctions. We perform positional encodings and identify key properties that make the positional encodings node embeddings. We found that the properties of expressiveness, efficiency and interpretability were achieved in the process. We saw that it is possible to use positional encodings as node embeddings, which can be used for machine learning tasks such as node classification, graph classification, and link prediction. We discuss some challenges and provide future directions.
93

Overlapping of Communication and Computation and Early Binding: Fundamental Mechanisms for Improving Parallel Performance on Clusters of Workstations

Dimitrov, Rossen Petkov 12 May 2001 (has links)
This study considers software techniques for improving performance on clusters of workstations and approaches for designing message-passing middleware that facilitate scalable, parallel processing. Early binding and overlapping of communication and computation are identified as fundamental approaches for improving parallel performance and scalability on clusters. Currently, cluster computers using the Message-Passing Interface for interprocess communication are the predominant choice for building high-performance computing facilities, which makes the findings of this work relevant to a wide audience from the areas of high-performance computing and parallel processing. The performance-enhancing techniques studied in this work are presently underutilized in practice because of the lack of adequate support by existing message-passing libraries and are also rarely considered by parallel algorithm designers. Furthermore, commonly accepted methods for performance analysis and evaluation of parallel systems omit these techniques and focus primarily on more obvious communication characteristics such as latency and bandwidth. This study provides a theoretical framework for describing early binding and overlapping of communication and computation in models for parallel programming. This framework defines four new performance metrics that facilitate new approaches for performance analysis of parallel systems and algorithms. This dissertation provides experimental data that validate the correctness and accuracy of the performance analysis based on the new framework. The theoretical results of this performance analysis can be used by designers of parallel system and application software for assessing the quality of their implementations and for predicting the effective performance benefits of early binding and overlapping. This work presents MPI/Pro, a new MPI implementation that is specifically optimized for clusters of workstations interconnected with high-speed networks. This MPI implementation emphasizes features such as persistent communication, asynchronous processing, low processor overhead, and independent message progress. These features are identified as critical for delivering maximum performance to applications. The experimental section of this dissertation demonstrates the capability of MPI/Pro to facilitate software techniques that result in significant application performance improvements. Specific demonstrations with Virtual Interface Architecture and TCP/IP over Ethernet are offered.
94

Investigation on Digital Fountain Codes over Erasure Channels and Additive White Gaussian Noise Channels

Huang, Weizheng 25 July 2012 (has links)
No description available.
95

Large-Message Nonblocking Allgather and Broadcast Offload via BlueField-2 DPU

Sarkauskas, Nicholas Robert 09 August 2022 (has links)
No description available.
96

Cumulative Distribution Networks: Inference, Estimation and Applications of Graphical Models for Cumulative Distribution Functions

Huang, Jim C. 01 March 2010 (has links)
This thesis presents a class of graphical models for directly representing the joint cumulative distribution function (CDF) of many random variables, called cumulative distribution networks (CDNs). Unlike graphical models for probability density and mass functions, in a CDN, the marginal probabilities for any subset of variables are obtained by computing limits of functions in the model. We will show that the conditional independence properties in a CDN are distinct from the conditional independence properties of directed, undirected and factor graph models, but include the conditional independence properties of bidirected graphical models. As a result, CDNs are a parameterization for bidirected models that allows us to represent complex statistical dependence relationships between observable variables. We will provide a method for constructing a factor graph model with additional latent variables for which graph separation of variables in the corresponding CDN implies conditional independence of the separated variables in both the CDN and in the factor graph with the latent variables marginalized out. This will then allow us to construct multivariate extreme value distributions for which both a CDN and a corresponding factor graph representation exist. In order to perform inference in such graphs, we describe the `derivative-sum-product' (DSP) message-passing algorithm where messages correspond to derivatives of the joint cumulative distribution function. We will then apply CDNs to the problem of learning to rank, or estimating parametric models for ranking, where CDNs provide a natural means with which to model multivariate probabilities over ordinal variables such as pairwise preferences. We will show that many previous probability models for rank data, such as the Bradley-Terry and Plackett-Luce models, can be viewed as particular types of CDN. Applications of CDNs will be described for the problems of ranking players in multiplayer team-based games, document retrieval and discovering regulatory sequences in computational biology using the above methods for inference and estimation of CDNs.
97

Decentralized network control, optimization and random walks on networks / Contrôle de réseau décentralisé, optimisation et marches aléatoires sur réseaux

De Bacco, Caterina 08 September 2015 (has links)
Dans les dernières années, plusieurs problèmes ont été étudiés à l'interface entre la physique statistique et l'informatique. La raison étant que, souvent, ces problèmes peuvent être réinterprétés dans le langage de la physique des systèmes désordonnés, où un grand nombre de variables interagit à travers champs locales qui dépendent de l'état du quartier environnant. Parmi les nombreuses applications de l'optimisation combinatoire le routage optimal sur les réseaux de communication est l'objet de la première partie de la thèse. Nous allons exploiter la méthode de la cavité pour formuler des algorithmes efficaces de type ‘’message-passing’’ et donc résoudre plusieurs variantes du problème grâce à sa mise en œuvre numérique. Dans un deuxième temps, nous allons décrire un modèle pour approcher la version dynamique de la méthode de la cavité, ce qui permet de diminuer la complexité du problème de l'exponentielle de polynôme dans le temps. Ceci sera obtenu en utilisant le formalisme de ‘’Matrix Product State’’ de la mécanique quantique.Un autre sujet qui a suscité beaucoup d'intérêt en physique statistique de processus dynamiques est la marche aléatoire sur les réseaux. La théorie a été développée depuis de nombreuses années dans le cas que la topologie dessous est un réseau de dimension d. Au contraire le cas des réseaux aléatoires a été abordé que dans la dernière décennie, laissant de nombreuses questions encore ouvertes pour obtenir des réponses. Démêler plusieurs aspects de ce thème fera l'objet de la deuxième partie de la thèse. En particulier, nous allons étudier le nombre moyen de sites distincts visités au cours d'une marche aléatoire et caractériser son comportement en fonction de la topologie du graphe. Enfin, nous allons aborder les événements rares statistiques associées aux marches aléatoires sur les réseaux en utilisant le ‘’Large deviations formalism’’. Deux types de transitions de phase dynamiques vont se poser à partir de simulations numériques. Nous allons conclure décrivant les principaux résultats d'une œuvre indépendante développée dans le cadre de la physique hors de l'équilibre. Un système résoluble en deux particules browniens entouré par un bain thermique sera étudiée fournissant des détails sur une interaction à médiation par du bain résultant de la présence du bain. / In the last years several problems been studied at the interface between statistical physics and computer science. The reason being that often these problems can be reinterpreted in the language of physics of disordered systems, where a big number of variables interacts through local fields dependent on the state of the surrounding neighborhood. Among the numerous applications of combinatorial optimisation the optimal routing on communication networks is the subject of the first part of the thesis. We will exploit the cavity method to formulate efficient algorithms of type message-passing and thus solve several variants of the problem through its numerical implementation. At a second stage, we will describe a model to approximate the dynamic version of the cavity method, which allows to decrease the complexity of the problem from exponential to polynomial in time. This will be obtained by using the Matrix Product State formalism of quantum mechanics. Another topic that has attracted much interest in statistical physics of dynamic processes is the random walk on networks. The theory has been developed since many years in the case the underneath topology is a d-dimensional lattice. On the contrary the case of random networks has been tackled only in the past decade, leaving many questions still open for answers. Unravelling several aspects of this topic will be the subject of the second part of the thesis. In particular we will study the average number of distinct sites visited during a random walk and characterize its behaviour as a function of the graph topology. Finally, we will address the rare events statistics associated to random walks on networks by using the large-deviations formalism. Two types of dynamic phase transitions will arise from numerical simulations, unveiling important aspects of these problems. We will conclude outlining the main results of an independent work developed in the context of out-of-equilibrium physics. A solvable system made of two Brownian particles surrounded by a thermal bath will be studied providing details about a bath-mediated interaction arising for the presence of the bath.
98

Development and application of an enhanced sampling molecular dynamics method to the conformational exploration of biologically relevant molecules

Alibay, Irfan January 2017 (has links)
This thesis describes the development a new swarm-enhanced sampling methodology and its application to the exploration of biologically relevant molecules. First, the development of a new multi-dimensional swarm-enhanced sampling molecular dynamics (msesMD) approach is detailed. Relative to the original swarm-enhanced sampling molecular dynamics (sesMD) methodology, the msesMD method demonstrates improved parameter transferability, resulting in more extensive sampling when scaling to larger systems such as alanine heptapeptide. The implementation and optimisation of the swarm-enhanced sampling algorithms in the AMBER software suite are also described. Through the use of the newer pmemd molecular dynamics (MD) engine and asynchronous MPI routines, speedups of up to three times the original sesMD implementation were achieved. The msesMD method is then applied to the investigation of carbohydrates, first looking at rare conformational changes in Lewis oligosaccharides. Validating against multi-microsecond unbiased MD trajectories and other enhanced sampling methods, the msesMD simulations identified rare conformational changes leading to the adoption of non-canonical unstacked core trisaccharide structures. Next, the use of msesMD as a tool to probe pyranose ring pucker events is explored. Evaluating against four benchmark monosaccharide systems, msesMD simulations accurately recover puckering details not easily obtained via multi-microsecond unbiased MD. This was followed by an exploration of the impact of ring substituents on conformation in glycosaminoglycan monosaccharides: through msesMD simulations, the influence of specific sulfation patterns were explored, finding that in some cases, such as 4-O-sulfation in N-acetyl-galactosamine, large changes in the relative stability of ring conformers can arise. Finally, the msesMD method was coupled with a thermodynamic integration scheme and used to evaluate solvation free energies for small molecule systems. Comparing against independent trajectory TI simulations, it was found that although the correct solvation free energies were obtained, the msesMD based method did not offer an advantage over unbiased MD for these small molecule systems. However, interesting discrepancies in free energy estimates arising from the use of hydrogen mass repartitioning were found.
99

Loose coupling and substitution principle in objet-oriented frameworks for web services / Couplage faible et principe de substitution dans les environnements à objets pour les services web

Allam, Diana 10 July 2014 (has links)
Actuellement, l’implémentation des services (modèles SOAP et RESTful) et de leurs applications clientes est de plus en plus basée sur la programmation par objet. Ainsi, les cadriciels orientés-objets pour les services Web sont essentiellement composés de deux couches : une couche à objets qui enveloppe une couche à services. Dans ce contexte, deux principes sont nécessaires pour la spécification de ces cadriciels : (i) En premier lieu, un couplage faible entre les deux couches, ce qui permet de cacher la complexité des détails techniques de la couche à services dans la couche à objets et de faire évoluer la couche à services avec un impact minimal sur la couche à objets (ii) En second lieu, une interopérabilité induite par le principe de substitution associée au sous-typage dans la couche à objets. Dans cette thèse, nous présentons d’abord les faiblesses existantes dans les cadriciels orientés-objets liés à ces deux principes. Ensuite, nous proposons une nouvelle spécification pour ces cadriciels en vue de résoudre ces problèmes. Comme application, nous décrivons la mise en œuvre de notre spécification dans le cadriciel cxf, à la fois pour SOAP et RESTful. / Today, the implementation of services (SOAP and RESTful models) and of client applications is increasingly based on object-oriented programming languages. Thus, object-oriented frameworks for Web services are essentially composed with two levels: an object level built over a service level. In this context, two properties could be particularly required in the specification of these frameworks: (i)First a loose coupling between the two levels, which allows the complex technical details of the service level to be hidden at the object level and the service level to be evolved with a minimal impact on the object level, (ii) Second, an interoperability induced by the substitution principle associated to subtyping in the object level, which allows to freely convert a value of a subtype into a supertype. In this thesis, first we present the existing weaknesses of object-oriented frameworks related to these two requirements. Then, we propose a new specification for object-oriented Web service frameworks in order to resolve these problems. As an application, we provide an implementation of our specification in the cxf framework, for both SOAP and RESTful models.
100

Métaheuristiques hybrides distribuées et massivement parallèles / Hybrid metaheuristics distributed and massively parallel

Abdelkafi, Omar 07 November 2016 (has links)
De nombreux problèmes d'optimisation propres à différents secteurs industriels et académiques (énergie, chimie, transport, etc.) nécessitent de concevoir des méthodes de plus en plus efficaces pour les résoudre. Afin de répondre à ces besoins, l'objectif de cette thèse est de développer une bibliothèque composée de plusieurs métaheuristiques hybrides distribuées et massivement parallèles. Dans un premier temps, nous avons étudié le problème du voyageur de commerce et sa résolution par la méthode colonie de fourmis afin de mettre en place les techniques d'hybridation et de parallélisation. Ensuite, deux autres problèmes d'optimisation ont été traités, à savoir, le problème d'affectation quadratique (QAP) et le problème de la résolution structurale des zéolithes (ZSP). Pour le QAP, plusieurs variantes basées sur une recherche taboue itérative avec des diversifications adaptatives ont été proposées. Le but de ces propositions est d'étudier l'impact de : l'échange des données, des stratégies de diversification et des méthodes de coopération. Notre meilleure variante est comparée à six des meilleurs travaux de la littérature. En ce qui concerne le ZSP, deux nouvelles formulations de la fonction objective sont proposées pour évaluer le potentiel des structures zéolitiques trouvées. Ces formulations sont basées sur le principe de pénalisation et de récompense. Deux algorithmes génétiques hybrides et parallèles sont proposés pour générer des structures zéolitiques stables. Nos algorithmes ont généré actuellement six topologies stables, parmi lesquelles trois ne sont pas répertoriées sur le site Web du SC-IZA ou dans l'Atlas of Prospective Zeolite Structures. / Many optimization problems specific to different industrial and academic sectors (energy, chemicals, transportation, etc.) require the development of more effective methods in resolving. To meet these needs, the aim of this thesis is to develop a library of several hybrid metaheuristics distributed and massively parallel. First, we studied the traveling salesman problem and its resolution by the ant colony method to establish hybridization and parallelization techniques. Two other optimization problems have been dealt, which are, the quadratic assignment problem (QAP) and the zeolite structure problem (ZSP). For the QAP, several variants based on an iterative tabu search with adaptive diversification have been proposed. The aim of these proposals is to study the impact of: the data exchange, the diversification strategies and the methods of cooperation. Our best variant is compared with six from the leading works of the literature. For the ZSP two new formulations of the objective function are proposed to evaluate the potential of the zeolites structures founded. These formulations are based on reward and penalty evaluation. Two hybrid and parallel genetic algorithms are proposed to generate stable zeolites structures. Our algorithms have now generated six stable topologies, three of them are not listed in the SC-JZA website or in the Atlas of Prospective Zeolite Structures.

Page generated in 0.1487 seconds