Avec le développement et la multiplication des appareils connectés dans tous les domaines, de nouvelles solutions pour le traitement de flux de données ont vu le jour. Cette thèse s'inscrit dans ce contexte: elle a été réalisée dans le cadre du projet FUI Waves, une plateforme de traitement de flux distribués. Le cas d'usage pour le développement a été la gestion des données provenant d'un réseau de distribution d'eau potable, plus précisément la détection d'anomalie dans les mesures de qualité et leur contextualisation par rapport à des données extérieures. Plusieurs contributions ont été réalisées et intégrées à différentes étapes du projet, leur évaluation et les publications liées témoignant de leur pertinence. Celles-ci se basent sur une ontologie que j'ai spécifiée depuis des échanges avec les experts du domaine travaillant dans chez le partenaire métier du projet. L'utilisation de données géographiques a permis de réaliser un système de profilage visant à améliorer le processus de contextualisation des erreurs. Un encodage de l'ontologie adapté au traitement de flux de données RDF a été développé pour supporter les inférences de RDFS enrichis de owl : sameAs. Conjointement, un formalisme compressé de représentation des flux (PatBin) a été conçu et implanté dans la plateforme. Il se base sur la régularité des motifs des flux entrants. Enfin, un langage de requêtage a été développé à partir de ce formalisme. Il intègre une stratégie de raisonnement se basant sur la matérialisation et la réécriture de requêtes. Enfin, à partir de déductions provenant d'un d'apprentissage automatique, un outil de génération de requêtes a été implanté. Ces différentes contributions ont été évaluées sur des jeux de données concrets du domaine ainsi que sur des jeux d'essais synthétiques / With the developpement and the expansion of connected devices in every domain, several projects on stream processing have been developped. This thesis has been realized as part of the FUI Waves, a reasoning stream processing engine distributed. The use case for the developement was the processing of data streamed from a potable water distribution network, more specifically the detection of anomalies in the quality measures and their contextualisation using external data. Several contributions have been realized and integrated in different stages of the project, wih evaluations and publications witnessing their relevance. These contributions use an ontology that has been designed thanks to collaboration with domain experts working for our water data management project partner. The use of geographical data allowed to realize a profiling system aiming at improving the anomaly contextualisation process. An ontology encoding approach, adapted to RDF stream processing, has been developped to support RDFS inferences enriched with owl : sameAs. Conjointly, a compressed formalism (PatBin) has been designed to represent streams. PatBin is based on the regularity of patterns found in incoming streams. Moreover, a query language has been conceived from PatBin, namely PatBinQL. It integrates a reasoning strategy that combines both materialization and query rewritting. Finally, given deductions coming from a Waves machine learning component, a query generation tool has been developped. These diferent contributions have been evaluated on both real-world and synthetic datasets

(1) Relational fusion of multiple features for the classical regression task (single measure and dimension). Originally, the task of recommender systems is formulated as a regression task. Many CF algorithms and fusion methods have been proposed. The limitation of previous fusion methods is that only local features are utilized and the global relational dependency is ignored, which would impair the performance of CF. We propose a relational fusion approach based on conditional random fields (CRF) to improve traditional fusion methods by incorporating global relational dependency. / (2) Fusion of regression-oriented and ranking-oriented algorithms for multi-measure adaption. Beyond the level of classical regression, ranking the items directly is another important task for recommender systems. A good algorithm should adapt to both regression-oriented and ranking-oriented measures. Traditionally, algorithms separately adapt to a single one, thus they cannot adapt to the other. We propose methods to combine them to improve the performances in both measures. / (3) Fusion of quality-based and relevance-based algorithms for multi-dimensional adaption. Recommender systems should consider the performances of multiple dimensions, such as quality and relevance. Traditional algorithms, however, only recommend either high-quality or high-relevance items. But they cannot adapt to the other dimension. We propose both fusion metrics and fusion approaches to effectively combine multiple dimensions for better performance in multi-dimensional recommendations. / (4) Investigation of impression efficiency optimization in recommendation. Besides performance, impression efficiency, which describes how much profit can be obtained per impression of recommendation, is also a very important issue. From recent study, over-quantity recommendation impression is intrusive to users. Thus the impression efficiency should be formulated and optimized. But this issue has rarely been investigated. We formulate the issue under the classical secretary problem framework and extend an online secretary algorithm to solve it. / Recommender systems are important nowadays. With the explosive growth of resources on the Web, users encounter information overload problem. The research issue of recommender systems is a kind of information filtering technique that suggests user-interested items (e.g., movies, books, products, etc.) to solve this problem. Collaborative filtering (CF) is the key approach. Over the decades, recommender systems have been demonstrated important in E-business. Thus designing accurate algorithms for recommender systems has attracted much attention. / This thesis is to investigate effective fusion-based approaches for recommender systems. Effective fusion of various features and algorithms becomes important along with the development of recommendation techniques. Because each feature/algorithm has its own advantages and disadvantages. A combination to get the best performance is desired in applications. The fusion-based approaches investigated are from the following four levels. / Xin, Xin. / Advisers: Wai Lam; Irwin Kuo Chin King; Michael Rung Tsong Lyu. / Source: Dissertation Abstracts International, Volume: 73-06, Section: B, page: . / Thesis (Ph.D.)--Chinese University of Hong Kong, 2011. / Includes bibliographical references (leaves 152-172). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [201-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese.

Thesis. 1975. M.S.--Massachusetts Institute of Technology. Dept. of Ocean Engineering. / Bibliography: leaves 234-235. / by John G. Lundblad. / M.S.

by John Pui-Fai Sum. / Thesis (Ph.D.)--Chinese University of Hong Kong, 1998. / Includes bibliographical references (p. 155-[163]). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Mode of access: World Wide Web.

In order to account for spatial correlation in residuals in regression models for areal and lattice data, different disciplines have developed distinct approaches. Bayesian spatial statistics typically has used a Gaussian conditional autoregressive (CAR) prior on random effects, while geographers utilize Moran's I statistic as a measure of spatial autocorrelation and the basis for creating spatial models. Recent work in both fields has recognized and built on a common feature of the two approaches, specifically the implicit or explicit incorporation into the linear predictor of eigenvectors of a matrix representing the spatial neighborhood structure. The inclusion of appropriate choices of these vectors effectively reduces the spatial autocorrelation found in the residuals. We begin with extensive simulation studies to compare Bayesian CAR models, Restricted Spatial Regression (RSR), Bayesian Spatial Filtering (BSF), and Eigenvector Spatial Filtering (ESF) with respect to estimation of fixed-effect coefficients, prediction, and reduction of residual spatial autocorrelation. The latter three models incorporate the neighborhood structure of the data through the eigenvectors of a Moran operator. We propose an alternative selection algorithm for all candidate predictors that avoids the ad hoc approach of RSR and selects on both model fit and reduction of autocorrelation in the residuals. The algorithm depends on the marginal posterior density a quantity that measures what proportion of the total variance can be explained by the measurement error. The algorithm selects candidate predictors that lead to a high probability that this quantity is large in addition to having large marginal posterior inclusion probabilities (PIP) according to model fit. Two methods were constructed. The first is based on orthogonalizing all of the candidate predictors while the second can be applied to the design matrix of candidate predictors without orthogonalization. Our algorithm was applied to the same simulated data that compared the RSR, BSF and ESF models. Although our algorithm performs similarly to the established methods, the first of our selection methods shows an improvement in execution time. In addition, our approach is a statistically sound, fully Bayesian method.

A fundamental question in visual neuroscience is how to represent image structure. The most common representational schemes rely on differential operators that compare adjacent image regions. While well-suited to encoding local relationships, such operators have significant drawbacks. Specifically, each filter's span is confounded with the size of its sub-fields, making it difficult to compare small regions across large distances. We find that such long-distance comparisons are more tolerant to common image transformations than purely local ones, suggesting they may provide a useful vocabulary for image encoding. . We introduce the "Dissociated Dipole," or "Sticks" operator, for encoding non-local image relationships. This operator de-couples filter span from sub-field size, enabling parametric movement between edge and region-based representation modes. We report on the perceptual plausibility of the operator, and the computational advantages of non-local encoding. Our results suggest that non-local encoding may be an effective scheme for representing image structure.

Thesis (Ph. D.)--Virginia Polytechnic Institute and State University, 1977. / Also available via the Internet.

History mechanisms available in hypertext systems allow users access to past interactions with the system and help users incorporate those interactions into the current context. The history information can be useful to both the system and the user. The Visual Knowledge Builder (VKB) creates spatial hypertexts - visual workspaces for collecting, organizing, and sharing. It is based on prior work on VIKI. VKB records all edit events and presents them in the form of a "navigable history" as end-users work within an information workspace. My thesis explores attaching user interpretations of history via the grouping and annotation of edit events. Annotations can take the form of a plain text statement or one or more attribute/value pairs attached to individual events or group of events in the list. Moreover, I explore the value of history event filtering, limiting the edits and groups presented to those that match user descriptions. My contribution in this thesis is the addition of mechanisms whereby users can cope with larger history records in VKB via the process of grouping, annotating and filtering history information.

A networked control system (NCS) is a feedback control system that has its control loop physically connected via real-time communication networks. To meet the demands of `teleautomation', modularity, integrated diagnostics, quick maintenance and decentralization of control, NCSs have received remarkable attention worldwide during the past decade. Yet despite their distinct advantages, NCSs are suffering from network-induced constraints such as time delays and packet dropouts, which may degrade system performance. Therefore, the network-induced constraints should be incorporated into the control design and related studies.<p> For the problem of state estimation in a network environment, we present the strategy of simultaneous input and state estimation to compensate for the effects of unknown input missing. A sub-optimal algorithm is proposed, and the stability properties are proven by analyzing the solution of a Riccati-like equation.<p> Despite its importance, system identification in a network environment has been studied poorly before. To identify the parameters of a system in a network environment, we modify the classical Kalman filter to obtain an algorithm that is capable of handling missing output data caused by the network medium. Convergence properties of the algorithm are established under the stochastic framework.<p> We further develop an adaptive control scheme for networked systems. By employing the proposed output estimator and parameter estimator, the designed adaptive control can track the expected signal. Rigorous convergence analysis of the scheme is performed under the stochastic framework as well.

This report describes the design and implementation of a fixed audio equalizer based on a scheme where parts of the signal spectrum are downsampled and treated differently for the purpose of reducing the computational complexity and memory requirements. The primary focus has been on finding a way of taking an equalizer based on a simple minimum-phase FIR filter and transform it to the new type of equalizer. To achieve this, a number of undesireable effects such as aliasing distortion and upsampling imaging had to be considered and dealt with. In order to achieve a good amplitude response of the system, optimization procedures were used. As part of the thesis, a cost-effective implementation of the filter has been made for an FPGA, in order to verify that the scheme is indeed usable for equalizing an audio signal.