Soil spectroscopy has experienced a tremendous increase in soil property characterisation, and can be used not only in the laboratory but also from the space (imaging spectroscopy). Partial least squares (PLS) regression is one of the most common approaches for the calibration of soil properties using soil spectra. Besides functioning as a calibration method, PLS can also be used as a dimension reduction tool, which has scarcely been studied in soil spectroscopy. PLS components retained from high-dimensional spectral data can further be explored with the gradient-boosted decision tree (GBDT) method. Three soil sample categories were extracted from the Land Use/Land Cover Area Frame Survey (LUCAS) soil library according to the type of land cover (woodland, grassland, and cropland). First, PLS regression and GBDT were separately applied to build the spectroscopic models for soil organic carbon (OC), total nitrogen content (N), and clay for each soil category. Then, PLS-derived components were used as input variables for the GBDT model. The results demonstrate that the combined PLS-GBDT approach has better performance than PLS or GBDT alone. The relative important variables for soil property estimation revealed by the proposed method demonstrated that the PLS method is a useful dimension reduction tool for soil spectra to retain target-related information.

Le modèle standard de la physique des particules, bien qu'étant une réussite incontestable n'explique toujours pas certaines observations, ce qui pourrais indiquer la présence d'une nouvelle physique. L'expérience ATLAS cherche à mettre en évidence ces nouveaux phénomènes en analysant les collisions de protons du LHC.Le prérequis pour toute analyse est d'avoir des données de bonne qualité. La première partie de cette thèse présente donc une étude sur les bouffées de bruit cohérent présentes dans les calorimètres à argon liquide d'ATLAS. Ce bruit a été traité pour garder un haut niveau de qualité des données.La deuxième partie de cette thèse se place dans le cadre de la recherche de nouvelle physique. Une résonance est cherchée dans le spectre en masse invariante des paires de quarks top. L'étude se concentre sur le cas des quarks top boostés se désintégrant de manière électronique. Cette topologie n'ayant jusqu'à présent pas été optimisée par la collaboration ATLAS, le travail de cette thèse a consisté à définir et développer une nouvelle méthode appelée "electron-in-jet removal" permettant d'améliorer la reconstruction des quarks top dans un tel régime.Cette méthode donne accès à des électrons qui auparavant étaient rejetés. Il faut donc mesurer les facteurs correctifs à appliquer pour corriger les imperfections de la simulation de ces électrons. Des mesures préliminaires de facteurs d'échelle d'identification sur les électrons se trouvant dans un jets ont ainsi été menées. Deux méthodes sont présentées ainsi qu'une réflexion sur les perspectives à apporter à ces mesures. / The standard model of particle physics is a very predictive theory, but it still fails to explain some observations and so leads to the idea of the existence of new physics. To discover it experimentally, the ATLAS collaboration analyses the proton-proton collisions provided by the LHC.Analyses need data of good quality. Hence, the first part of this document describes a work to characterize the coherent noise bursts observed in the liquid argon calorimeters of the ATLAS experiment. Such noise has been studied and cured to ensure a high level of data quality.The second part of this thesis takes place in the context of a new physics search using top quark pairs. This study is focused on the case where boosted top quarks decay electronically. The reconstruction of top quarks in such cases was never optimized by the ATLAS collaboration. Therefore this study has lead to the definition and implementation of a new method called "electron-in-jet removal" improving significantly the top quark reconstruction in such topologies.This new method gives access to electrons that were previously removed. The correction factor applied to correct the imperfections of the simulation should then be computed for those electrons. Some preliminary measurements have been performed for the cases where the electron is within a jet. Two methods are presented as well as thoughts about future implementations.

No estudo das comunidades florestais, estabelecer a importância relativa dos fatores que definem a composição e a distribuição das espécies é um desafio. Em termos de gradientes ambientais o estudo das respostas das espécies arbóreas são essenciais para a compreensão dos processos ecológicos e decisões de conservação. Neste sentido, para contribuir com a elucidação dos processos ecológicos nas principais formações florestais do Estado de São Paulo (Floresta Ombrófila Densa de Terras Baixas, Floresta Ombrófila Densa Submontana, Floresta Estacional Semidecidual e Savana Florestada) este trabalho objetivou responder as seguintes questões: (I) a composição florística e a abundância das espécies arbóreas, em cada unidade fitogeográfica, variam conforme o gradiente edáfico e topográfico?; (II) características do solo e topografia podem influenciar na previsibilidade de ocorrência de espécies arbóreas de ampla distribuição em diferentes tipos vegetacionais? (III) existe relação entre o padrão de distribuição espacial de espécies arbóreas e os parâmetros do solo e topografia? O trabalho foi realizado em parcelas alocadas em unidades de conservação (UC) que apresentaram trechos representativos, em termos de conservação e tamanho, das quatro principais formações florestais presentes no Estado de São Paulo. Em cada UC foram contabilizados os indivíduos arbóreos (CAP ≥ 15 cm), topografia, dados de textura e atributos químicos dos solos em uma parcela de 10,24 ha, subdividida em 256 subparcelas. Análises de correspodência canônica foram aplicadas para estabelecer a correspondência entre a abundância das espécies e o gradiente ambiental (solo e topografia). O método TWINSPAN modificado foi aplicado ao diagrama de ordenação da CCA para avaliar a influência das variáveis ambientais (solo e topografia) na composição de espécies. Árvores de regressão \"ampliadas\" (BRT) foram ajustadas para a predição da ocorrência das espécies segundo as variáveis de solo e topografia. O índice de Getis-Ord (G) foi utilizado para determinar a autocorrelação espacial das variáveis ambientais utilizadas nos modelos de predição da ocorrência das espécies. Nas unidades fitogeográficas analisadas, a correspondência entre o gradiente ambiental (solo e topografia) e a abundância das espécies foi significativa, especialmente na Savana Florestada onde observou-se a maior relação. O solo e a topografia também se relacionaram com a semelhança na composição florística das subparcelas, com exceção da Floresta Estacional Semicidual (EEC). As principais variáveis de solo e topografia relacionadas a flora em cada UC foram: (1) Na Floresta Ombrófila Densa de Terras Baixas (PEIC) - teor de alumínio na camada profunda (Al (80-100 cm)) que pode refletir os teor de Al na superfície, acidez do solo (pH(H2O) (5-25 cm)) e altitude, que delimitou as áreas alagadas; (2) Na Floresta Ombrófila Densa Submontana (PECB) - altitude, fator que, devido ao relevo acidentado, influencia a temperatura e incidência de sol no sub-bosque; (3) Na Savana Florestada (EEA) - fertilidade, tolerância ao alumínio e acidez do solo. Nos modelos de predição BRT, as variáveis químicas dos solos foram mais importantes do que a textura, devido à pequena variação deste atributo no solo nas áreas amostradas. Dentre as variáveis químicas dos solos, a capacidade de troca catiônica foi utilizada para prever a ocorrência das espécies nas quatro formações florestais, sendo particularmente importante na camada mais profunda do solo da Floresta Ombrófila Densa de Terras Baixas (PEIC). Quanto à topografia, a altitude foi inserida na maioria dos modelos e apresentou diferentes influências sobre as áreas de estudo. De modo geral, para presença das espécies de ampla distribuição observou-se uma mesma tendência quando à associação com os atributos dos solos, porém com amplitudes dos descritores edáficos que variaram de acordo com a área de estudo. A ocorrência de Guapira opposita e Syagrus romanzoffiana, cujo padrão variou conforme a escala, foi explicada por variáveis com padrões espaciais agregados que somaram entre 30% e 50% de importância relativa no modelo BRT. A presença de A. anthelmia, cujo padrão também apresentou certo nível de agregação, foi associada apenas a uma variável com padrão agregado, a altitude (21%), que pode ter exercido grande influência na distribuição da espécie ao delimitar áreas alagadas. T. guianensis se associou a variáveis ambientais preditoras com padrão espacial agregado que somaram cerca de 70% de importância relativa, o que deve ter sido suficiente para estabelecer o padrão agregado em todas as escalas. No entanto, a influência dos fatores ambientais no padrão de distribuição da espécie não depende apenas do ótimo ambiental da espécie, mas um resultado da interação espécie-ambiente. Concluiu-se que: (I) características edáficas e topográficas explicaram uma pequena parcela da composição florística, em cada unidade fitogeográfica, embora a ocorrência de algumas espécies tenha se associado ao gradiente edáfico e topográfico; (II) a partir de características dos solos e da topografia foi possível prever a presença de espécies arbóreas, que apresentaram particularidades em relação a sua associação com o solo de cada fitofisionomia; (III) a partir de associações descritivas o solo e a topografia influenciam o padrão de distribuição espacial das espécies, na proporção em que contribuem para a presença das mesmas. / In the study of forest communities, establish the relative importance of the factors that define the composition and distribution of species is a challenge. In terms of environmental gradients study the responses of tree species are essential to the understanding of ecological processes and conservation decisions. In this regard, to contribute to the elucidation of ecological processes in the main forest formations of São Paulo (Dense Ombrophylous Forest of Lowlands, Submontane Dense Ombrophylous Forest, Semideciduous Forest and Savanna Woodland) this study aimed to answer the following questions: (I) floristic composition and tree species abundance in each phytogeographic unit change according to edaphic and topographic gradient?; (II) soil characteristics and topography can influence the occurrence of predictability of tree species widely distributed in different types of vegetation? (III) there is a relationship between spatial distribution pattern of tree species and the soil parameters and topography? The work was carried out in allocated plots in protected areas (PA) with the four main forest formations in terms of conservation and size of Sao Paulo. In each PA was sampled individual trees, topography, texture data and chemical properties of the soil on a plot of 10.24 ha, subdivided into 256 subplots. Canonical corresponding analyzes (CCA) were applied to establish the correspondence between the abundance of species and environmental gradient (soil and topography). The modified TWINSPAN method was applied to CCA ordination diagram to evaluate the influence of environmental variables (soil and topography) on species composition. Boosteed Regression Trees (BRT) were adjusted for predicting the occurrence of the species according to soil variables and topography. The Getis Ord-index (G) was used to determine the spatial autocorrelation of environmental variables used in the BRT models. In analyzed phytogeographic units, correspondence between the environmental gradient (soil and topography) and abundance of species was significant, especially in Savanna Woodland. The soil and topography also correlated with the floristic composition similarity of the subplots, with the exception of Semicidual Seasonal Forest (EEC). The main soil and topography variables related to floristic in each PA were: (1) Dense Ombrophylous Forest of Lowlands (PEIC) - aluminium content in the deep layer (Al (80-100 cm)) which may reflect the Al content at the surface, soil acidity (pH (H2O) (5-25 cm)) and altitude, which outlined the flooded areas; (2) Submontane Dense Ombrophylous Forest (PECB) - elevation, due to the rugged terrain influences the temperature and light incidence in the understory; (3) Savanna Woodland (EEA) - fertility, tolerance to aluminum and soil acidity. In BRT prediction models, the chemical soil variables were more important than the texture due to small variation of this soil attribute in the sampled area. Among the soil chemical variables, cation exchange capacity was used to predict the species occurrence in four forest formations and particularly important in the soil deepest layer on the Dense Ombrophylous Forest of Lowlands (PEIC). In relation to topography, elevation was included in most models and had different influences on the study areas. Overall, the species widely distributed showed the same trend as the association with the attributes of the soil, but with amplitudes of edaphic descriptors that change according to the study area. The occurrence of the Guapira opposita and Syagrus romanzoffiana, whose pattern change according to the scale, was explained by variables with aggregated spatial patterns that amounted to between 30% and 50% relative importance in the BRT model. The presence of A. anthelmia, which defaults also presented certain level of aggregation, was associated only with one aggregate variable, elevation (21%), which may have exerted great influence on the species distribution to delimit wetlands. T. guianensis was related with the predictive environmental variables of aggregate spatial pattern which totaled to about 70% relative importance, what must have been enough to establish the aggregate pattern at all scales. However, the influence of environmental factors (soil and topography) on the species distribution pattern depends not only on the environmental optimum of the species, but a result of species-environment interaction. We concluded that: (I) soil and topographical characteristics explain a small portion of the floristic composition in each phytogeographic unit, although the occurrence of some species have been associated to the soil and topographic gradient; (II) from soil characteristics and topography it was possible to predict the presence of tree species, which showed particular in relation to its association with the soil of each vegetation type; (III) from descriptive associations soil and topography influence the spatial distribution pattern of the species, to the extent that contribute to the presence of the same.

The Highway Safety Manual (HSM) estimates roadway safety performance based on predictive models that were calibrated using national data. Calibration factors are then used to adjust these predictive models to local conditions for local applications. The HSM recommends that local calibration factors be estimated using 30 to 50 randomly selected sites that experienced at least a total of 100 crashes per year. It also recommends that the factors be updated every two to three years, preferably on an annual basis. However, these recommendations are primarily based on expert opinions rather than data-driven research findings. Furthermore, most agencies do not have data for many of the input variables recommended in the HSM. This dissertation is aimed at determining the best way to meet three major data needs affecting the estimation of calibration factors: (1) the required minimum sample sizes for different roadway facilities, (2) the required frequency for calibration factor updates, and (3) the influential variables affecting calibration factors. In this dissertation, statewide segment and intersection data were first collected for most of the HSM recommended calibration variables using a Google Maps application. In addition, eight years (2005-2012) of traffic and crash data were retrieved from existing databases from the Florida Department of Transportation. With these data, the effect of sample size criterion on calibration factor estimates was first studied using a sensitivity analysis. The results showed that the minimum sample sizes not only vary across different roadway facilities, but they are also significantly higher than those recommended in the HSM. In addition, results from paired sample t-tests showed that calibration factors in Florida need to be updated annually. To identify influential variables affecting the calibration factors for roadway segments, the variables were prioritized by combining the results from three different methods: negative binomial regression, random forests, and boosted regression trees. Only a few variables were found to explain most of the variation in the crash data. Traffic volume was consistently found to be the most influential. In addition, roadside object density, major and minor commercial driveway densities, and minor residential driveway density were also identified as influential variables.

Découvert en 1995 à Fermilab, le quark top est le dernier quark découvert. La mesure de ses propriétés permet de tester les prédictions du Modèle Standard et de contraindre la masse du boson de Higgs. De par ses propriétés, le quark top est aussi un partenaire privilégié dans la recherche de particules de Nouvelle Physique attendues à l'échelle du TeV. Ce travail de thèse, effectué auprès du détecteur ATLAS au LHC, présente les méthodes mises en oeuvre afin de mesurer la masse du quark top dans sa désintégration semileptonique. Différentes méthodes de reconstruction des évènements top sont présentées ainsi qu'une analyse dédiée basée sur les arbres de décision boostés. Ses performances sont quantifiées. La mesure précise de la masse du quark top nécessite une compréhension approfondie de l'échelle en énergie des jets. Deux stratégies sont présentées afin de calibrer les jets légers et les jets issus de quark b à l'échelle partonique. Les performances d'un ajustement cinématique appliqué à la mesure de la masse du quark top sont présentées. Une mesure de la masse du quark top est extraite en utilisant une définition de la masse calibrée à l'échelle partonique. / Discovered in 1995 at Fermilab, top quark is the last quark discovered. The measurement of its properties allows to test Standard Model predictions and to constraint Higgs boson mass. Due to its properties, the top quark is a privileged partner in the search for New Physics particles expected around TeV scale. This thesis, performed using the ATLAS detector at LHC, describes the different methods developed in order to measure precisely the top quark mass in its semileptonic decay. Two reconstruction methods are presented as well as a dedicated one based on Boosted Decision Trees. Its performances are quantified The precise measurement of the top quark mass needs a deep understanding of the jet energy scale. This thesis presents two strategies to calibrate light and b jets to the partonic scale. The performance of a kinematical fit applied to top mass measurement are presented. A precise measurement of the top quark mass is done using a calibrated scale to the partonic level.

Soil spectroscopy has experienced a tremendous increase in soil property characterisation, and can be used not only in the laboratory but also from the space (imaging spectroscopy). Partial least squares (PLS) regression is one of the most common approaches for the calibration of soil properties using soil spectra. Besides functioning as a calibration method, PLS can also be used as a dimension reduction tool, which has scarcely been studied in soil spectroscopy. PLS components retained from high-dimensional spectral data can further be explored with the gradient-boosted decision tree (GBDT) method. Three soil sample categories were extracted from the Land Use/Land Cover Area Frame Survey (LUCAS) soil library according to the type of land cover (woodland, grassland, and cropland). First, PLS regression and GBDT were separately applied to build the spectroscopic models for soil organic carbon (OC), total nitrogen content (N), and clay for each soil category. Then, PLS-derived components were used as input variables for the GBDT model. The results demonstrate that the combined PLS-GBDT approach has better performance than PLS or GBDT alone. The relative important variables for soil property estimation revealed by the proposed method demonstrated that the PLS method is a useful dimension reduction tool for soil spectra to retain target-related information.

Depuis la découverte du boson de Higgs en 2012, les physiciens et physiciennes des particules tentent de trouver des signes de nouvelle physique. Bien que le modèle standard décrivant les forces et particules aient été confirmé par les expériences telles que ATLAS au Grand collisionneur de hadrons (LHC), ce modèle décrit seulement 5% de la matière de notre Univers. Face à l’absence d’excès dans les recherches concentrées sur des modèles de nouvelle physique, l’intelligence artificielle pourrait être la solution. Ce mémoire s’inscrit dans une perspective novatrice de recherche générale de nouvelle physique indépendante des modèles théoriques dans les données du détecteur ATLAS, par l’utilisation de l’apprentissage machine. Ici, nous nous intéressons à l’application de réseaux de neurones dans les données de simulation de collision proton-proton à \sqrt{s} = 13 TeV du détecteur ATLAS. Nous étudierons la performance d’auto-encodeurs variationnels dans les jets boostés, ces jets qui pourraient cacher des signes de nouvelle physique. Pour analyser la performance de notre réseau, nous entraînons celui-ci sur les quadri-vecteurs de jets issus de gluons et de quarks légers. Nous tentons de retrouver un signal test, les jets issus de quarks top boostés, dans les données de simulation en effectuant des sélections sur les scores d’anomalie retournés par le réseau. Nos modèles d’auto-encodeurs variationnels atteignent une bonne discrimination entre le bruit de fond et le signal du quark top. Nous devrons toutefois améliorer le rejet du bruit de fond pour purifier notre signal en fonction de nos sélections. / Since the discovery of the Higgs boson in 2012, particle physicists are searching for signs of new physics. Although the standard model describing forces and particles has been confirmed by experiments like ATLAS at the Large Hadron Collider (LHC), it only describes 5% of the matter of the universe. Facing the absence of excess in searches for new physics model, artificial intelligence could be the solution. This master thesis is part of a novel general model-independent search for new physics in the ATLAS detector data using machine learning. Here, we are interested in the application of neural networks in \sqrt{s} = 13 TeV proton-proton collision ATLAS simulation data. We study the performance of variational auto-encoders in boosted jets, who might be hiding signs of new physics. To analyze our network performance, we train the network on jets four-vectors coming from gluons and light quarks. We try to find a test signal, boosted top quark jets, in our simulation data by applying selections on the anomaly scores returned by our network. Our variational auto-encoder reach a good discrimination between the background and the top quark signal. However, we need to improve background rejection to purify our signal as a function of our selections.

Due to the intermittent nature of renewable energy production, maintaining the stability of the power supply system is becoming a significant challenge of the energy transition. Besides, the penetration of Electric Vehicles (EVs) and the development of a large network of charging stations will inevitably increase the pressure on the electrical grid. However, this network and the batteries that are connected to it also constitute a significant resource to provide ancillary services and therefore a new opportunity to stabilize the power grid. This requires to be able to produce accurate short term forecasts of the power consumption of charging stations at distribution scale. This work proposes a full forecasting framework, from the transformation of discrete charging sessions logs into a continuous aggregated load profile, to the pre-processing of the time series and the generation of predictions. This framework is used to identify the most appropriate model to provide two days ahead predictions of the hourly load profile of large charging stations networks. Using three years of data collected at Amsterdam’s public stations, the performance of several state-of-the-art forecasting models, including Gradient Boosted Trees (GBTs) and Recurrent Neural Networks (RNNs) is evaluated and compared to a classical time series model (Auto Regressive Integrated Moving Average (ARIMA)). The best performances are obtained with an Extreme Gradient Boosting (XGBoost) model using harmonic terms, past consumption values, calendar information and temperature forecasts as prediction features. This study also highlights periodical patterns in charging behaviors, as well as strong calendar effects and an influence of temperature on EV usage. / På grund av den intermittenta karaktären av förnybar energiproduktion, blir upprätthållandet av elnäts stabilitet en betydande utmaning. Dessutom kommer penetrationen av elbilar och utvecklingen av ett stort nät av laddstationer att öka trycket på elnätet. Men detta laddnät och batterierna som är anslutna till det utgör också en betydande resurs för att tillhandahålla kompletterande tjänster och därför en ny möjlighet att stabilisera elnätet. För att göra sådant bör man kunna producera korrekta kortsiktiga prognoser för laddstationens strömförbrukning i distributions skala. Detta arbete föreslår ett fullständigt prognos protokoll, från omvandlingen av diskreta laddnings sessioner till en kontinuerlig förbrukningsprofil, till förbehandling av tidsserier och generering av förutsägelser. Protokollet används för att identifiera den mest lämpliga metoden för att ge två dagars förutsägelser av timförbrukning profilen för ett stort laddstation nät. Med hjälp av tre års data som samlats in på Amsterdams publika stationer utvärderas prestanda för flera avancerade prognosmodeller som är gradient boosting och återkommande neurala nätverk, och jämförs med en klassisk tidsseriemodell (ARIMA). De bästa resultaten uppnås med en XGBoost modell med harmoniska termer, tidigare förbrukningsvärden, kalenderinformation och temperatur prognoser som förutsägelse funktioner. Denna studie belyser också periodiska mönster i laddningsbeteenden, liksom starka kalendereffekter och temperaturpåverkan på elbilar-användning.

No description available.

No description available.