Analytic and force-directed placement methods that simultaneously minimize wire length and spread cells are receiving renewed attention from both academia and industry. However, these methods are by no means trivial to implement---to date, published works have failed to provide sufficient engineering details to replicate results. This dissertation addresses the implementation of a generic force-directed placer entitled FDP. Specifically, this thesis provides (1) a description of efficient force computation for spreading cells, (2) an illustration of numerical instability in this method and a means to avoid the instability, (3) metrics for measuring cell distribution throughout the placement area, and (4) a complementary technique that aids in minimizing wire length. FDP is compared to Kraftwerk and other leading academic tools including Capo, Dragon, and mPG for both standard cell and mixed-size circuits. Wire lengths produced by FDP are found to be, on average, up to 9% and 3% better than Kraftwerk and Capo, respectively. All told, this thesis confirms the validity and applicability of the approach, and provides clarifying details of the intricacies surrounding the implementation of a force-directed global placer.

Perfluoroalkyl acids (PFAAs) are man-made chemicals which have been observed in the global environment, even in locations far away from where they are emitted. These persistent substances are taken up in humans and biota and may have toxic effects. Knowledge about how PFAAs are dispersed in the environment is needed to discern strategies to manage their sources and to evaluate the efficacy of adopted legislation. This thesis aimed to increase our understanding of the sources of PFAAs to the atmosphere and how PFAAs are transported in air. The results of Paper I demonstrated that gaseous perfluorooctanoic acid (PFOA) sorbs to typical glass fibre filters (GFFs) used in high-volume air sampling of PFAAs. As a consequence, the fraction of gaseous PFOA present in sampled air is underestimated, while the fraction of PFOA associated with aerosols is overestimated. Replacing GFFs with filters deactivated through silanisation and siliconisation did not eliminate this sampling artefact and is therefore not recommended as a means to determine the gas-particle partitioning of PFAAs. In Paper II, monitoring of the mass of PFOA transferred from water solutions of pH 0.2-5.5 demonstrated that the acid dissociation constant of linear PFOA and the four most ubiquitous branched PFOA isomers is around or below 1. Furthermore, the results demonstrated that the presence of counter ions and organic matter in water retarded, rather than enhanced, the volatilisation of PFOA. Therefore, volatilisation of all isomers of PFOA from environmental waters is expected to be negligible. To further study the transfer of PFAAs from environmental waters to air, Paper III simulated the process of sea spray generation in the laboratory. Strong enrichment of PFAAs was observed from bulk water to the surface microlayer and to aerosols. The enrichment increased with PFAA chain length, indicating that this process is of greater importance for more surface active substances. The highest enrichment was observed in aerosols &lt; 1.6 µm, which can travel over long distances if not rained out. Based on the measured aerosol enrichment factors we estimated that approximately 70 metric tonnes of PFAAs are aerosolised from the global oceans yearly and that 3% of this mass is deposited in terrestrial environments. Paper IV reported the occurrence of branched PFOA isomers in deposition sampled in five geographical locations. The presence of these isomers demonstrated that atmospheric transformation of fluorotelomer alcohols is not the only ongoing source of PFAAs to air. We hypothesised that, additionally, both sea spray aerosols and direct emissions from manufacturing sources contributed to the contamination of the precipitation on different spatial scales. Although further research is required to determine the relative importance of different sources to the atmosphere locally and globally, this thesis has substantially advanced the state-of-the-science by i) demonstrating the significance of an air sampling artefact discussed as an uncertainty in the scientific literature over the past decade, ii) definitively ruling out volatilisation from environmental waters as a source of PFOA to air, iii) demonstrating transfer of PFAAs from seawater to air via sea spray aerosols and thus quantifying the environmental importance of this process, and iv) ultimately demonstrating that several types of sources of PFAAs impact the global atmosphere and thus PFAA contamination patterns in precipitation. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 3: Manuscript. Paper 4: Manuscript.</p>

No description available.

The WikiLeaks Afghanistan war logs contain more than 76 000 reports about fatalities and their circumstances in the US led Afghanistan war, covering the period from January 2004 to December 2009. In this paper we use those reports to build statistical models to help us understand the mortality rates associated with specific circumstances. We choose an approach that combines Latent Dirichlet Allocation (LDA) with negative binomial based recursive partitioning. LDA is used to process the natural language information contained in each report summary. We estimate latent topics and assign each report to one of them. These topics - in addition to other variables in the data set - subsequently serve as explanatory variables for modeling the number of fatalities of the civilian population, ISAF Forces, Anti-Coalition Forces and the Afghan National Police or military as well as the combined number of fatalities. Modeling is carried out with manifest mixtures of negative binomial distributions estimated with model-based recursive partitioning. For each group of fatalities, we identify segments with different mortality rates that correspond to a small number of topics and other explanatory variables as well as their interactions. Furthermore, we carve out the similarities between segments and connect them to stories that have been covered in the media. This provides an unprecedented description of the war in Afghanistan covered by the war logs. Additionally, our approach can serve as an example as to how modern statistical methods may lead to extra insight if applied to problems of data journalism. (author's abstract) / Series: Research Report Series / Department of Statistics and Mathematics

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal.

Durant cette dernière décennie, la diffusion du format XML pour représenter les données générées par et échangées sur le Web a été accompagnée par la mise en œuvre de nombreux moteurs d’évaluation de requêtes et de mises à jour XQuery. Parmi ces moteurs, les systèmes « mémoire centrale » (Main-memory Systems) jouent un rôle très important dans de nombreuses applications. La gestion et l’intégration de ces systèmes dans des environnements de programmation sont très faciles. Cependant, ces systèmes ont des problèmes de passage à l’échelle puisqu’ils requièrent le chargement complet des documents en mémoire centrale avant traitement.Cette thèse présente une technique de partitionnement des documents XML qui permet aux moteurs « mémoire principale » d’évaluer des expressions XQuery (requêtes et mises à jour) pour des documents de très grandes tailles. Cette méthode de partitionnement s’applique à une classe de requêtes et mises à jour pertinentes et fréquentes, dites requêtes et mises à jour itératives.Cette thèse propose une technique d'analyse statique pour reconnaître les expressions « itératives ». Cette analyse statique est basée sur l’extraction de chemins à partir de l'expression XQuery, sans utilisation d'information supplémentaire sur le schéma. Des algorithmes sont spécifiés, utilisant les chemins extraits par l’étape précédente, pour partitionner les documents en entrée en plusieurs parties, de sorte que la requête ou la mise à jour peut être évaluée sur chaque partie séparément afin de calculer le résultat final par simple concaténation des résultats obtenus pour chaque partie. Ces algorithmes sont mis en œuvre en « streaming » et leur efficacité est validée expérimentalement.En plus, cette méthode de partitionnement est caractérisée également par le fait qu'elle peut être facilement implémentée en utilisant le paradigme MapReduce, permettant ainsi d'évaluer une requête ou une mise à jour en parallèle sur les données partitionnées. / With the widespread diffusion of XML as a format for representing data generated and exchanged over the Web, main query and update engines have been designed and implemented in the last decade. A kind of engines that are playing a crucial role in many applications are « main-memory » systems, which distinguish for the fact that they are easy to manage and to integrate in a programming environment. On the other hand, main-memory systems have scalability issues, as they load the entire document in main-memory before processing. This Thesis presents an XML partitioning technique that allows main-memory engines to process a class of XQuery expressions (queries and updates), that we dub « iterative », on arbitrarily large input documents. We provide a static analysis technique to recognize these expressions. The static analysis is based on paths extracted from the expression and does not need additional schema information. We provide algorithms using path information for partitioning the input documents, so that the query or update can be separately evaluated on each part in order to compute the final result. These algorithms admit a streaming implementation, whose effectiveness is experimentally validated. Besides enabling scalability, our approach is also characterized by the fact that it is easily implementable into a MapReduce framework, thus enabling parallel query/update evaluation on the partitioned data.

Recursive partitioning algorithms separate a feature space into a set of disjoint rectangles. Then, usually, a constant in every partition is fitted. While this is a simple and intuitive approach, it may still lack interpretability as to how a specific relationship between dependent and independent variables may look. Or it may be that a certain model is assumed or of interest and there is a number of candidate variables that may non-linearly give rise to different model parameter values. We present an approach that combines generalized linear models with recursive partitioning that offers enhanced interpretability of classical trees as well as providing an explorative way to assess a candidate variable's in uence on a parametric model. This method conducts recursive partitioning of a generalized linear model by (1) fitting the model to the data set, (2) testing for parameter instability over a set of partitioning variables, (3) splitting the data set with respect to the variable associated with the highest instability. The outcome is a tree where each terminal node is associated with a generalized linear model. We will show the method's versatility and suitability to gain additional insight into the relationship of dependent and independent variables by two examples, modelling voting behaviour and a failure model for debt amortization, and compare it to alternative approaches.

Cette thèse propose une modélisation des interactions entre le Transport de Marchandises en Ville (TMV) et la dynamique du trafic. Ces deux composantes, pourtant imbriquées, ne sont que très rarement étudiées simultanément. Plus précisément, cette thèse investigue les impacts du TMV sur les états de trafic ; et inversement, les impacts des états de trafic sur le TMV. Les recherches s’articulent autour de deux axes : le TMV à l’échelle (i) de la livraison et (ii) de la tournée de livraison, et la dynamique du trafic à l’échelle (iii) d’un jour et (iv) d’un catalogue de jours. (i) Le TMV réalisé par des véhicules routiers peut engendrer des formes de congestion, et par conséquent des nuisances sonores et atmosphériques. Néanmoins, il s’agit d’un « mal nécessaire » car le transport de ces marchandises est essentiel pour l’économie des villes. Une pratique courante est la livraison en double file. Ces « stationnements sauvages » réduisent l’écoulement des véhicules dans le réseau de transport. Une étude de sensibilité quantitative en simulation microscopique met en lumière une dégradation significative des états de trafic sur un boulevard urbain à partir d’un faible nombre de stationnements en double file. Ces phénomènes locaux et temporaires semblent essentiels à intégrer dans les simulations du trafic afin d’étudier au mieux les performances globales du système de transport. (ii) Les tournées de livraison planifiées peuvent être améliorées par l’intégration des conditions de circulation. Les algorithmes existants d’optimisation de tournées se basent sur une certaine qualité et quantité de données. Nous investiguons l’influence de plusieurs niveaux de granularité d’information du trafic sur l’ordonnancement optimale des tournées. De plus, nous comparons les temps de parcours estimés par les algorithmes, et les temps de parcours effectifs estimés à l’aide d’une simulation microscopique. Les états de trafic attendus peuvent être sujets à quelques variations. Nous proposons une méthodologie de génération de tournées intégrant de l’incertitude dans la dynamique du trafic. (iii) À l’échelle de la ville, les états de trafic sont naturellement variables, que ce soit dans l’espace ou au cours du temps. Le partitionnement permet de découper un réseau de transport en zones homogènes. Une zone est définie par un ensemble de liens connectés ayant des conditions de circulation similaires. Nous proposons une approche spatio-temporelle définissant des zones en 3 Dimensions (3D). L’idée est de résumer la majeure partie de la dynamique du trafic d’une ville en utilisant peu d’information : la vitesse moyenne par zone spatio-temporelle. Deux familles de méthodes de clustering fondamentalement différentes sont comparées et évaluées. Le cas d’étude est le réseau d’Amsterdam avec des données réelles. (iv) D’un jour à l’autre, la dynamique du trafic peut être similaire. Nous proposons une méthodologie regroupant les jours par leurs motifs de congestion. L’existence d’une régularité dans les motifs journaliers est introduite par la notion de jour-type. Un jour-type est le jour de référence d’un groupe de jours. Nous validons notre modèle en comparant les temps de parcours effectifs et les temps de parcours estimés par les jours-types. Diverses applications peuvent être raffinées à partir de quelques jours-types, comme l’assistance routière et la génération de tournées. / This thesis aims to modeling the interactions between the urban freight and the traffic dynamic. Although both are intertwined, they have rarely been studied considering a combined approach. Specifically, the main purpose is to analyze the impacts of the urban freight on traffic states, and vice versa. This research is focused on two axes: the urban freight considering (i) the delivery and (ii) the delivery route scales, and two traffic dynamic scales describing (iii) a day and (iv) a set of days. (i) The delivery of goods carried by vehicles can produce traffic congestion, noise and air pollution. Nevertheless, the transport of goods is essential for the economy of cities. The double parking is a common way for delivery trucks. This illegal parking reduces the vehicle capacity of the transportation network. A sensitivity analysis in traffic microscopic simulation model for urban freight highlight a significant decreasing of the traffic conditions on an urban corridor considering a low number of double parking. These local and temporal phenomena are essential to be integrated into the traffic simulation in order to better study the overall performance of the transportation system. (ii) The delivery route planner can be improved by integrating traffic conditions. The route optimization algorithms are based on quality and quantity of given data available. We investigate the influence of several levels of granularity on traffic information data for the optimal route scheduling. Moreover, the travel time estimated by algorithms and the effective travel time estimated by a microscopic simulation are compared. In addition, the expected traffic conditions can be subject to variations. We propose a methodology of delivery route planner integrating the uncertainty of the traffic dynamics. (iii) At the city level, the traffic conditions are varied through space and time. A partitioning of urban transportation networks makes possible to identify homogeneous zones. A zone is defined by a set of connected links with similar traffic conditions. We proposed a spatial and temporal approach to define the 3D zones. The idea is to summarize the majority of the traffic dynamics of a given city using only a few information: the mean speed per spatial and temporal zone. Two fundamentally different methods of clustering are compared and evaluated. The study case is the Amsterdam network with its real-world traffic data. (iv) From day to day, the traffic dynamic can be similar. We proposed a methodology grouping days by their similar congestion patterns. The existence of a regularity through daily patterns are introduced by consensual speed map. A consensual speed map is the reference day representing a group of days. Our model is validated by using the effective travel time and estimated ones by the consensual speed maps. Numerous applications can be refined from a couple of consensual speed maps, as the route guidance and delivery route planning.

Antimoine (Sb) et tellure (Te), sont des contaminants peu étudiés (isotopes stables) et leurs radionucléides artificiels peuvent être rejetés dans le milieu aquatique lors des accidents nucléaires. La connaissance de leurs comportements biogéochimiques respectifs est nécessaire à l'évaluation du risque radiologique post-accidentel.Ce travail présente des données originales sur le comportement biogéochimique de Sb et de Te dans les systèmes de transition continent-océan, tels que l'estuaire de la Gironde et la rivière du Rhône. Un suivi de 14 ans et des campagnes océanographiques dans le bassin versant de l’estuaire de la Gironde ont permis d’identifier des concentrations, des flux, et des réactivités (variabilités spatio-temporelles et distribution solide/liquide) plus élevés pour Sb que pour Te, mettant en évidence un comportement additif pour Sb et de soustraction pour Te le long des gradients de salinité et de turbidité estuariennes. Des expériences couplant l’adsorption d’isotopes marqués sur des matières en suspension (MES) et des extractions sélectives des phases porteuses, suggèrent que les formes apportées de Sb et de Te sont plus mobiles et potentiellement plus biodisponibles que leurs équivalents naturels. De plus, l’observation de la bioaccumulation non-négligeable de Sb et de Te naturels dans les huîtres sauvages à l’embouchure de l’estuaire permet d’envisager une absorption potentielle de leurs homologues radioactifs.Ainsi, le développement de scenarios de dispersion de radionucléides rejetés dans les zones de transition dépendra (i) de la position géographique de la source (Rhône) et/ou de la zone de turbidité maximale (ZTM; système fluvio-estuarien de Gironde), (ii) de la situation hydrologique pendant et post accident, ainsi que (iii) de la réactivité biogéochimique et des temps de demi-vies des radionucléides. Les premiers scénarios de dispersion de radionucléides dans l'estuaire de la Gironde suggèrent (i) un transport préférentiel de Sb dissous vers la zone côtière, et (ii) une forte rétention de Te radioactif dans la ZTM si la dernière est présente en aval du site d’accident, impliquant le risque de migration saisonnière de la radioactivité vers la ville de Bordeaux pendant l’étiage suivant. Ainsi, la dynamique intra estuarienne (marée, débit et migration de la ZTM) sera le facteur prédominant dans le devenir de Te radioactif, depuis son rejet jusqu’à sa désintégration complète en iode radioactif. L’ensemble de ce travail met en évidence la nécessité d’une évaluation plus approfondie de la radiotoxicité potentielle de Sb et Te lors de leurs rejets en milieu aquatique. / Antimony (Sb) and tellurium (Te) are relatively uncommon contaminants (stable isotopes) and may form short-lived fission products (radionuclides) released into the environment during nuclear power plants accidents. Little is known about their respective biogeochemical behaviours, necessary for general contamination studies and post-accidental radiological risk assessment.This work provides original knowledge on Sb and Te biogeochemical behaviour in highly dynamic continent-ocean transition systems: the Gironde Estuary and the Rhône River. Concentrations, spatial/temporal variations, solid/liquid partitioning (Kd), and fluxes are studied from long-term records at the watershed scale. Four estuarine sampling campaigns during contrasting hydrological conditions show higher Sb solubility and Te particle affinity in the estuary than in the upstream fluvial reaches. Historical records (1984-2017) in wild oysters from the estuary mouth do not show clear trends of past or recent contamination, but measurable bioaccumulation suggests that potential uptake of radionuclides is likely to occur. Combined adsorption experiments using isotopically-labelled (spiked) Sb and Te, and subsequent selective extractions of carrier phases from suspended particulate matter (SPM) suggest that spiked Sb and Te are more mobile and potentially bioaccessible than their environmental (inherited) equivalents. Radiotracer adsorption experiments using environmentally representative concentrations of both Gironde and Rhône systems underpin that highly soluble elements may show contrasting reactivity between inherited and spiked forms.Radionuclide dispersion will greatly depend on (i) the geographical position of the source (Rhône) and/or the maximum turbidity zone (MTZ; Gironde fluvial-estuarine system), (ii) the succession of hydrological situations during and after the accident, and (iii) the biogeochemical reactivity and half-lives of the radionuclides. First scenarios on hypothetical dissolved radionuclide dispersion in the Gironde Estuary suggest (i) low sorption of Sb to the SPM, implying a transport of radionuclides in dissolved phase towards the coast, and (ii) high retention of Te within the MTZ, especially for accidental releases during flood conditions, linking the fate of radioactive Te to long estuarine SPM residence times (1-2 years). Potential upstream migration of Te radionuclides in the MTZ towards the city of Bordeaux during the following summer season and Te decay into radioactive iodine warrants further evaluation of the associated potential radiotoxicity.

Purpose – The purpose of this thesis is to understand how specialist firms develop in a concentrated industry by analyzing how microbreweries develop and operate on the Belgian market. Then, the literature reviewed used to build the theoretical foundation of this work is centered around the Resource Partitioning Theory and Niche Marketing Theory to explain this phenomenon. Design/methodology/approach – A cross-case analysis was undertaken. Five Belgian Microbreweries were examined. The data were gathered through interviews conducted with owners of microbreweries based on a structured questionnaire, as well as through secondary data. In addition, an analysis of the Belgian beer industry was conducted, using as secondary data, such as the existing literature on the beer industry, as well annual reports from Beers Associations. Findings – In opposition to the big companies that dominates the center of the market leveraging on economies of scales, the microbreweries strived to create unique products that have a high quality in term of taste. Moreover, they focused their efforts in their region, creating value first for local consumers and not all have an ambition to sell in the whole country or to internationalize the product. Thus, the microbreweries managed to create both functional and symbolic value to the consumers. However, the authors found that the symbolic value created by the breweries had more component than cultural heritage, local and specialist identity. Then, relationship was of importance for the microbreweries as they sought personal contact with their consumer and tried to be involved in their community as well as cooperated with other breweries (coopetition).   Overall, the microbreweries practices followed what is described in niche-marketing and entrepreneurial literature. Originality/value – First, this research participates to the resource partitioning literature by contributing to the limited knowledge on how specialist firms are operating, Then, this exploratory study is focused on an understudied area that is how microbreweries runs, that is especially true on the Belgian market as it was barely explored before. Thereafter, this thesis can be used as inspiration for new researches in different countries or with different approaches and/or theories.