Les oscillations acoustiques de baryons (BAO) constituent une règle standard permettant de contraindre les différents modèles cosmologiques. Cette thèse rend compte de la première mesure des BAO dans la fonction de corrélation de la fraction de flux transmise des forêts Lyman-α des quasars à grands décalages spectraux. Cette détection utilise 89322 spectres de quasars mesurés par le Baryon Oscillation Spectroscopic Survey (BOSS) de la troisième génération du Sloan Digital Sky Survey (SDSS-III). Les quasars considérés possèdent des décalages spectraux compris dans l'intervalle 2,1 < z < 3,5. Un pic dans la fonction de corrélation est détecté à 1,043+0,021−0,020 fois la position attendu du pic BAO pour le modèle de concordance ΛCDM. Cette mesure permet de contraindre la distance angulaire DA ainsi que le paramètre de Hubble H à un décalage spectral moyen z = 2,38. Par ailleurs cette thèse présente une nouvelle méthode de sélection des quasars par variabilité. Cette méthode est appliquée à la région du Stripe 82 où un grand nombre de données photométriques multi- époque est disponible. Sur cette région, elle permet d'atteindre une densité d'environ 30 deg−2 quasars contre 18 deg−2 pour les sélections usuelles par couleur.

In the wake of the Big Data phenomenon, the computing world has seen a number of computational paradigms developed in response to the sudden need to process ever-increasing volumes of data. Most notably, MapReduce has proven quite successful in scaling out an extensible class of simple algorithms to even hundreds of thousands of nodes. However, there are some tasks---even embarrassingly parallelizable ones---that neither MapReduce nor any existing automated parallelization framework is well-equipped to perform. For instance, any computation that (naively) requires consideration of all pairs of inputs becomes prohibitively expensive even when parallelized over a large number of worker nodes. Many of the most desirable methods in machine learning and statistics exhibit these kinds of all-pairs or, more generally, all-tuples computations; accordingly, their application in the Big Data setting may seem beyond hope. However, a new algorithmic strategy inspired by breakthroughs in computational physics has shown great promise for a wide class of computations dubbed generalized N-body problems (GNBPs). This strategy, which involves the simultaneous traversal of multiple space-partitioning trees, has been applied to a succession of well-known learning methods, accelerating each asymptotically and by orders of magnitude. Examples of these include all-k-nearest-neighbors search, k-nearest-neighbors classification, k-means clustering, EM for mixtures of Gaussians, kernel density estimation, kernel discriminant analysis, kernel machines, particle filters, the n-point correlation, and many others. For each of these problems, no overall faster algorithms are known. Further, these dual- and multi-tree algorithms compute either exact results or approximations to within specified error bounds, a rarity amongst fast methods. This dissertation aims to unify a family of GNBPs under a common framework in order to ease implementation and future study. We start by formalizing the problem class and then describe a general algorithm, the generalized fast multipole method (GFMM), capable of solving all problems that fit the class, though with varying degrees of speedup. We then show O(N) and O(log N) theoretical run-time bounds that may be obtained under certain conditions. As a corollary, we derive the tightest known general-dimensional run-time bounds for exact all-nearest-neighbors and several approximated kernel summations. Next, we implement a number of these algorithms in a commercial database, empirically demonstrating dramatic asymptotic speedup over their conventional SQL implementations. Lastly, we implement a fast, parallelized algorithm for kernel discriminant analysis and apply it to a large dataset (40 million points in 4D) from the Sloan Digital Sky Survey, identifying approximately one million quasars with high accuracy. This exceeds the previous largest catalog of quasars in size by a factor of ten and has since been used in a follow-up study to confirm the existence of dark energy.

Magister Scientiae - MSc / We investigate the modelling of radio galaxies within a semi-analytic framework in the Millennium Simulation of the Virgo Consortium. The aim is to assess the radio sources contamination of Sunyaev-Zeldovich (SZ) signatures of clusters of galaxies in Cosmic Microwave Background (CMB) experiments. The modelling is also relevant to the Karoo Array Telescope (MeerKAT) and the Square Kilometre Array (SKA) science. The semi-analytical model consists of N-body simulation, the Millennium Run to trace the merger history of dark matter haloes within the Λ Cold Dark Matter (ΛCDM) cosmology and a follow up of the black hole accretion history and Active Galactic Nuclei (AGN) evolution. We study the growth of the supermassive black hole (SMBH) in galaxy centres and determine the black hole mass accretion conversion into radiation. We identify a model which matches observed radio luminosity function. We describe a model of observed sample of radio surveys at a given frequency and a flux density limit to obtain a model of radio luminosity function (space density of radio sources as a function of redshift) that we compare with our simulated data. We determine the redshift distribution of radio galaxies (FRI), blazars and radio quasars (FRII) in the simulation. We focus the modelling on flat spectrum population of blazars since their jets are collimated towards us and thus constitute the most potential contaminants of the CMB. We determine the spatial and density distribution of radio sources in clusters with a virial mass Mvir 2 1014h−1M and then compute the temperature fluctuations and fluxes produced by these cluster radio sources. Our main results include: the model provides a reasonable match within uncertainties with the model obtained by Dunlop & Peacock (1990) [39] using their best fit of radio luminosity function at redshift z . 0:3. The model underestimates the number of radio sources at high redshift z & 1. Radio sources are concentrated around the centre of clusters with a maximum density at r . 0:1r200 where r200 is the radius within which the density is 200 times the critical density. Radio sources are more concentrated in low mass clusters. The model predicts a surface density profile of radio sources with luminosity P 1023 W.Hz−1 at 1.4 GHz (z . 0:06) in agreement with that of Lin & Mohr (2007) [58] at r . 0:1r200 but underestimates the density in the outskirts of the clusters. BL Lacs and FRI radio galaxies produce non negligible contamination at redshift z . 0:1. They produce a mean temperature fluctuation 4:5 K at redshift z 0:01 which can be at the same level as the kinetic SZE signal produced by the cluster. Blazars constitute potential contaminant of the thermal SZ effect at redshift z 1:0 and z 1:5 at 145 GHz where they produce a mean temperature 300 K - 350 K for an average mass of the cluster. / South Africa

No description available.

No description available.