Jurnal Edukasi
Not a member yet
4879 research outputs found
Sort by
L'éponge de Menger -itération 4-
No full textThe Menger sponge -iteration 4- (L'éponge de Menger -itération 4-
Private and Public Provision of Counseling to Job-Seekers: Evidence from a Large Controlled Experiment
No full textThis paper reports the results of a large scaled randomized controlled experiment comparing the public and private provision of counseling to job-seekers. The intention-to-treat estimates of both programs are not statistically different, but more workers were enrolled in the private program, implying an effect per beneficiary that is twice as large under the public as under the private program. We find suggestive evidence that the private firms may have insu fficiently mastered the counseling technology, and exercised less effort on those who had the best chance to find a job. This highlights the incentive problems in designing contracts for these services
Accelerating Metropolis-Hastings algorithms: Delayed acceptance with prefetching
No full text20 pages, 12 figures, 2 tables, submittedMCMC algorithms such as Metropolis-Hastings algorithms are slowed down by the computation of complex target distributions as exemplified by huge datasets. We offer in this paper an approach to reduce the computational costs of such algorithms by a simple and universal divide-and-conquer strategy. The idea behind the generic acceleration is to divide the acceptance step into several parts, aiming at a major reduction in computing time that outranks the corresponding reduction in acceptance probability. The division decomposes the "prior x likelihood" term into a product such that some of its components are much cheaper to compute than others. Each of the components can be sequentially compared with a uniform variate, the first rejection signalling that the proposed value is considered no further, This approach can in turn be accelerated as part of a prefetching algorithm taking advantage of the parallel abilities of the computer at hand. We illustrate those accelerating features on a series of toy and realistic examples
Deux applications arithmétiques des travaux d'Arthur
No full textWe present two arithmetic applications of James Arthur's endoscopic classification of the discrete automorphic spectrum for symplectic and orthogonal groups. The first one consists in removing the irreducibility assumption in a theorem of Richard Taylor describing the image of complex conjugations by p-adic Galois representations associated with regular, algebraic, essentially self-dual, cuspidal automorphic representations of GL_{2n+1} over a totally real number field. We also extend it to the case of representations of GL_{2n} whose multiplicative character is ''odd''. We use a p-adic deformation argument, more precisely we prove that on the eigenvarieties for symplectic and even orthogonal groups, there are ''many'' points corresponding to (quasi-)irreducible Galois representations. Arthur's endoscopic classification is used to define these Galois representations, and also to transfer self-dual automorphic representations of the general linear group to these classical groups. The second application concerns the explicit computation of dimensions of spaces of automorphic or modular forms. Our main contribution is an algorithm computing orbital integrals at torsion elements of an unramified p-adic classical group, for the unit of the unramified Hecke algebra. It allows to compute the geometric side in Arthur's trace formula, and thus the Euler characteristic of the discrete spectrum in level one. Arthur's endoscopic classification allows to analyse precisely this Euler characteristic, and deduce the dimensions of spaces of level one automorphic forms. The dimensions of spaces of vector-valued Siegel modular forms, which constitute a more classical problem, are easily derived.Nous proposons deux applications à l'arithmétique des travaux récents de James Arthur sur la classification endoscopique du spectre discret des groupes symplectiques et orthogonaux. La première consiste à ôter une hypothèse d'irréductibilité dans un résultat de Richard Taylor décrivant l'image des conjugaisons complexes par les représentations galoisiennes p-adiques associées aux représentations automorphes cuspidales algébriques régulières essentiellement autoduales pour le groupe GL_{2n+1} sur un corps totalement réel. Nous l'étendons également au cas de GL_{2n}, sous une hypothèse de parité du caractère multiplicatif. Nous utilisons un résultat de déformation p-adique. Plus précisément, nous montrons l'abondance de points correspondant à des représentations galoisiennes (quasi-)irréductibles sur les variétés de Hecke pour les groupes symplectiques et orthogonaux pairs. La classification d'Arthur est utilisée à la fois pour définir les représentations galoisiennes et pour transférer des représentations automorphes autoduales (pas nécessairement cuspidales) de groupes linéaires aux groupes symplectiques et orthogonaux. La deuxième application concerne le calcul explicite de dimensions d'espaces de formes automorphes ou modulaires. Notre contribution principale est un algorithme calculant les intégrales orbitales aux éléments de torsion des groupes classiques p-adiques non ramifiés, pour l'unité de l'algèbre de Hecke non ramifiée. Cela permet le calcul du côté géométrique de la formule des traces d'Arthur, et donc celui de la caractéristique d'Euler du spectre discret en niveau un. La classification d'Arthur permet l'analyse fine de cette caractéristique d'Euler, jusqu'à en déduire les dimensions des espaces de formes automorphes. De là il n'est pas difficile d'apporter une réponse à un problème plus classique: déterminer les dimensions des espaces de formes modulaires de Siegel à valeurs vectorielles
Application of the arbitrary decomposition to finite spot size Mueller matrix measurements
No full textInternational audienceFinite spot size Mueller matrix polarimetric measurements whereby the light spot impinges on two different areas of the sample, e.g., a grating and a substrate, are relatively frequently met in practice. It has been shown that if the Mueller matrix of one of the areas (the substrate) is known from an additional measurement then the Mueller matrix of the remaining medium (the grating) can be obtained from the (substrate-grating) overall response by the polarimetric subtraction method. We show that, provided a specific condition is fulfilled, the individual polarimetric responses of the two areas can be retrieved from the finite spot size measurement by using a special form of the arbitrary decomposition even if none of the individual responses is known a priori. The decomposition method is illustrated on a microelectronics grating structure and its accuracy, as well as limits of applicability, is discusse
Quantum Dynamics of the Driven and Dissipative Rabi Model
No full text13 pages, final versionInternational audienceThe Rabi model considers a two-level system (or spin-1/2) coupled to a quantized harmonic oscillator and describes the simplest interaction between matter and light. The recent experimental progress in solid-state circuit quantum electrodynamics has engendered theoretical efforts to quantitatively describe the mathematical and physical aspects of the light-matter interaction beyond the rotating wave approximation. We develop a stochastic Schr\"{o}dinger equation approach which enables us to access the strong-coupling limit of the Rabi model and study the effects of dissipation, and AC drive in an exact manner. We include the effect of ohmic noise on the non-Markovian spin dynamics resulting in Kondo-type correlations, as well as cavity losses. We compute the time evolution of spin variables in various conditions. As a consideration for future work, we discuss the possibility to reach a steady state with one polariton in realistic experimental conditions
Market composition and price informativeness in a large market with endogenous order types
No full textWe analyse the joint determination of price informativeness and the composition of the market by order type in a large asset market with dispersed information. The market microstructure is one in which informed traders may place market orders or full demand schedules and where market makers set the price. Market-order traders trade less aggressively on their information and thus reduce the informativeness of the price; in a full market-order market, price informativeness is bounded, whatever the quality of tradersinformation about the assets dividend. When traders can choose their order type and demand schedules are (even marginally) costlier than market orders, then market-order traders overwhelm the market when the precision of private signals goes to innity. This is because demand schedules are substitutes: at high levels of precision, a residual fraction of demand-schedule traders is sufficient to take the trading price close to traders signals, while the latter is itself well aligned with the dividend. Hence, the gain from trading conditional on the price (as demand-schedule traders do) in addition to ones own signal (as all informed traders do) vanishes
Evolution of genes neighborhood within reconciled phylogenies: an ensemble approach
No full textInternational audienceWe consider a recently introduced dynamic programming scheme to compute parsimonious evolutionary scenarios for gene adjacencies. We extend this scheme to sample evolutionary scenarios from the whole solution space under the Boltzmann distribution. We apply our algorithms to a dataset of mammalian gene trees and adjacencies, and observe a significant reduction of the number of syntenic inconsistencies observed in the resulting ancestral gene adjacencies
Marangoni induced force on a drop in a Hele Shaw cell
No full textInternational audienceWe analyse the force balance on a cylindrical drop in a Hele-Shaw cell, subjected to a Marangoni flow caused by a surface tension gradient. Depth-averaged Stokes equations, called Brinkman equations, are introduced and a general closed form solution is obtained. The validity of the averaging procedure is ascertained by considering a linear surface tension gradient acting on a cylindrical flattened drop. The Marangoni-driven flow field and resulting force predicted by the Brinkman model are seen to match well a full three-dimensional direct numerical simulation. A closed form ex-pression of the force acting on the drop is obtained, calculated from contributions due to the normal viscous stress, tangential viscous stress, and pressure fields, integrated on the drop perimeter. This expression is used to predict the force balance when a stationary droplet is submitted to both a carrier flow and a Marangoni flow. We show that previous results in the literature had underestimated by a factor two the Marangoni-induced force