La raison et son Dieu. Etude sur la théologie kantienneTheis, Robert ![]() Book published by Librarie philosophique J. Vrin (2012) Detailed reference viewed: 36 (0 UL) Raisonnement scientifique et inférence bayésienneBernard, Nicolas ![]() Bachelor/master dissertation (2013) Après deux siècles de quasi-oubli, les méthodes dites bayésiennes ont connu un essor extraordinaire au cours des trente dernières années. Ce sont de telles applications de la théorie des probabilités qui ... [more ▼] Après deux siècles de quasi-oubli, les méthodes dites bayésiennes ont connu un essor extraordinaire au cours des trente dernières années. Ce sont de telles applications de la théorie des probabilités qui faisaient dire de celle-ci au principal pionnier de ces méthodes, Laplace, qu'elle «n'est que le bon sens réduit au calcul». Des formalisations plus récentes (Cox, 1946) font que l'on sait aujourd'hui que raisonner de manière rationnelle est équivalent à utiliser un raisonnement bayésien [Jaynes 2003]. Savoir raisonner de manière rationnelle étant nécessaire pour tout scientifique, il nous a semblé intéressant de faire le lien entre ces méthodes bayésiennes et ce que l'on appelle généralement la méthode scientifique. Les méthodes bayésiennes sont des méthodes qui dérivent de l'application de l'interprétation bayésienne de la théorie des probabilités. Selon cette interprétation, une probabilité n'est pas une propriété intrinsèque d'un système (comme une fréquence), mais une mesure (qui peut être subjective) de l'incertitude que l'on a sur l'état de celui-ci. Le coeur de ces méthodes est alors le théorème de Bayes (entre 1745 et 1750), qui permet de réviser cette mesure à l'aune de nouvelles données: p(H|D,I) = p(D|H,I) * p(H|I)/p(D|I). La notation utilisée ici reflète l'usage qui nous intéresse: H représente une hypothèse, D des données, et, un raisonnement ne se faisant pas «dans le vide», I représente un contexte informationnel qui n'intervient pas dans les calculs mais que nous avons choisi de représenter explicitement, suivant ainsi [Sivia 2006]. Le terme p(H|I) représente alors la probabilité qui avait été attribuée a priori à l'hypothèse, p(D|I) représente les faits nouveaux que l'on désire prendre en compte, et p(D|H,I) la vraisemblance de ces faits en supposant l'hypothèse vraie. On extrait de cela la probabilité révisée, a posteriori de l'hypothèse, p(H|D,I). Dans cette étude, nous expliquons que l'application de ce théorème est non seulement compatible avec la méthode scientifique traditionnelle venant de la tradition d'Aristote, Bacon et Descartes et ses trois étapes --observer, former des hypothèses, tester ces hypothèses-- mais qu'en plus, elle permet d'étendre la logique «cartésienne» pour prendre en compte le problème de l'inférence. En effet, la logique cartésienne permet de déduire les conséquences à partir de causes connues, mais il n'est pas possible de déterminer formellement les causes à partir de l'observation des faits. L'utilisation du théorème de Bayes, autrefois appelé loi des probabilités inverses, permet de résoudre ce problème, en déterminant la plausibilité de diverses hypothèses à partir des faits. En outre, le remplacement d'une logique binaire avec deux valeurs «vrai/faux» par un continuum de probabilité permet «d'adoucir» l'effet d'observations contraires à une théorie par ailleurs largement validée, alors qu'une application brutale et naïve de la logique classique voudrait qu'elle soit considérée immédiatement comme réfutée. Inversement, l'application du théorème de Bayes avec des données prédites par une théorie confirme l'idée intuitive que cette théorie est rendue plus plausible (sans devenir pour autant certaine). Ce sont des avantages par rapport au réfutationnisme poppérien. Cette extension de la logique facilite également la prise en compte de données dont l'interprétation est incertaine. Les méthodes bayésiennes fournissent aussi un critère pour résoudre le problème qui peut se poser avec le principe du rasoir d'Occam. En effet, celui-ci veut qu'entre deux théories à pouvoir explicatif identique, on choisisse celle qui est la plus simple. Mais que faire dans le cas où s'affrontent deux théories, dont l'une est plus complexe mais aussi plus précise que l'autre~? Le théorème de Bayes permet de comparer les probabilités respectives de ces théories en prenant non seulement en compte leur adéquation aux données dont on dispose, mais également le nombre de paramètres (donc la complexité) prévu par ces théories. Faute d'avoir le temps de présenter cela à l'aide d'un exemple comme nous l'avions initialement espéré, nous renvoyons le lecteur vers le chapitre 4 de [Sivia-2006]. Incidemment, le formalisme bayésien de l'induction de Solomonoff permet même une démonstration du rasoir d'Occam. De manière plus générale, étant donné qu'il n'est guère possible de rendre compte de l'incroyable richesse des applications des méthodes bayésiennes, nous dirigeons le lecteur soucieux d'en apprendre plus vers notre bibliographie annotée. [less ▲] Detailed reference viewed: 268 (15 UL) Les raisons de la peur: Les classes moyennes sont-elles protégées de la crise ?Chauvel, Louis ![]() in Revue de l'OFCE : Observations et Diagnostics Economiques (2012), 18 Detailed reference viewed: 151 (9 UL) La “Raison” de l’éducation, le “social” et l’étude de la scolarisationPopkewitz, Thomas S. ![]() in Dossier: La raison educative comme question sociale (2013) Detailed reference viewed: 30 (0 UL) The Ramakrishna-Taylor method and modular lifts of mod p^n Galois representationsAdibhatla, Rajender ![]() Presentation (2013, June 21) Detailed reference viewed: 49 (0 UL) Raman analysis of monoclinic Cu2SnS3 thin films; Djemour, Rabie ; Gütay, Levent et alin Applied Physics Letters (2012), 100 Detailed reference viewed: 222 (6 UL) Raman imaging of doping domains in graphene on SiO(2); ; et al in Applied Physics Letters (2007), 91(24), We present spatially resolved Raman images of the G and 2D lines of single-layer graphene flakes. The spatial fluctuations of G and 2D lines are correlated and are thus shown to be affiliated with local ... [more ▼] We present spatially resolved Raman images of the G and 2D lines of single-layer graphene flakes. The spatial fluctuations of G and 2D lines are correlated and are thus shown to be affiliated with local doping domains. We investigate the position of the 2D line-the most significant Raman peak to identify single-layer graphene-as a function of charging up to vertical bar n vertical bar approximate to 4x10(12) cm(-2). Contrary to the G line which exhibits a strong and symmetric stiffening with respect to electron and hole doping, the 2D line shows a weak and slightly asymmetric stiffening for low doping. Additionally, the linewidth of the 2D line is, in contrast to the G line, doping independent making this quantity a reliable measure for identifying single-layer graphene. (C) 2007 American Institute of Physics. [less ▲] Detailed reference viewed: 85 (0 UL) Raman imaging of graphene; ; et al in Solid State Communications (2007), 143(1-2), 44-46 A Raman spectrum of a solid contains information about its vibrational and electronic properties. Collecting spectral data with spatial resolution and encoding it in a 2D plot generates images with ... [more ▼] A Raman spectrum of a solid contains information about its vibrational and electronic properties. Collecting spectral data with spatial resolution and encoding it in a 2D plot generates images with information complementary to optical and scanning force imaging. In the case of few-layer graphene the frequency of the G line and especially the width of the D ' line turn out to be sensitive to single layers. The thickness of the few-layer graphene flake is reflected in the intensity of the G line and in the reduced intensity of the dominant peak of the underlying silicon oxide. (c) 2007 Elsevier Ltd. All rights reserved. [less ▲] Detailed reference viewed: 56 (1 UL) Raman mapping of a single-layer to double-layer graphene transition; ; et al in European Physical Journal. Special Topics (2007), 148 We report on confocal Raman spectroscopy on a few-layer graphene flake. Adjacent single- and double-layer graphene sections allow mapping the transition in vibrational and electronic properties to a ... [more ▼] We report on confocal Raman spectroscopy on a few-layer graphene flake. Adjacent single- and double-layer graphene sections allow mapping the transition in vibrational and electronic properties to a second stacked graphene sheet and with it a weak interlayer coupling. Most prominently the width of the D' peak doubles upon going from a single to a double layer, which can be explained within the double-resonant Raman model. The intensities of the G and G' lines decrease at the crossover to a single layer. Contrary to the G' line the G peak position shifts to higher wave numbers, however, not uniformly over the entire section: its frequency fluctuates spatially. The Raman map of the D line intensity shows a non-zero contribution at the boundaries of the flake and the individual sections, which can be attributed either to defects and disorder or to the breakdown of translational symmetry, whereas within the flake no D line signal is detected. [less ▲] Detailed reference viewed: 134 (0 UL) Raman signatures of ferroic domain walls captured by principal component analysis; ; Kreisel, Jens et alin Journal of Physics: Condensed Matter (2018), 30(3), Ferroic domain walls are currently investigated by several state-of-the art techniques in order to get a better understanding of their distinct functional properties. Here, principal component analysis ... [more ▼] Ferroic domain walls are currently investigated by several state-of-the art techniques in order to get a better understanding of their distinct functional properties. Here, principal component analysis (PCA) of Raman maps is used to study ferroelectric domain walls (DWs) in LiNbO3 and ferroelastic DWs in NdGaO3. It is shown that PCA allows us to quickly and reliably identify small Raman peak variations at ferroelectric DWs and that the value of a peak shift can be deduced-accurately and without a priori-from a first order Taylor expansion of the spectra. The ability of PCA to separate the contribution of ferroelastic domains and DWs to Raman spectra is emphasized. More generally, our results provide a novel route for the statistical analysis of any property mapped across a DW. [less ▲] Detailed reference viewed: 84 (3 UL) Raman spectra of BN nanotubes: Ab initio and bond-polarizability model calculationsWirtz, Ludger ; ; et alin Physical Review. B, Condensed Matter and Materials Physics (2005), 71(24), We present ab initio calculations of the nonresonant Raman spectra of zigzag and armchair BN nanotubes. In comparison, we implement a generalized bond-polarizability model where the parameters are ... [more ▼] We present ab initio calculations of the nonresonant Raman spectra of zigzag and armchair BN nanotubes. In comparison, we implement a generalized bond-polarizability model where the parameters are extracted from first-principles calculations of the polarizability tensor of a BN sheet. For light polarization along the tube axis, the agreement between model and ab initio spectra is almost perfect. For perpendicular polarization, depolarization effects have to be included in the model in order to reproduce the ab initio Raman intensities. [less ▲] Detailed reference viewed: 110 (0 UL) Raman spectroscopy as probe of nanometre-scale strain variations in graphene; Reichardt, Sven ; et alin Nature Communications (2015), 6 Detailed reference viewed: 218 (7 UL) Raman spectroscopy as probe of nanometre-scale strain variations in graphene; Reichardt, Sven ; et alPoster (2015, July 14) Detailed reference viewed: 112 (1 UL) Raman Spectroscopy of GrapheneReichardt, Sven ; Wirtz, Ludger ![]() in Binder, Rolf (Ed.) Optical Properties of Graphene (2017) Raman spectroscopy of graphene is reviewed from a theoretical perspective. After an introduction of the building blocks (electronic band structure, phonon dispersion, electron-phonon interaction, electron ... [more ▼] Raman spectroscopy of graphene is reviewed from a theoretical perspective. After an introduction of the building blocks (electronic band structure, phonon dispersion, electron-phonon interaction, electron-light coupling), Raman intensities are calculated using time-dependent perturbation theory. The analysis of the contributing terms allows for an intuitive understanding of the Raman peak positions and intensities. The Raman spectrum of pure graphene only displays two principle peaks. Yet, their variation as a function of internal and external parameters and the occur- rence of secondary, defect-related peaks, conveys a lot of information about the system. Thus, Raman spectroscopy is used routinely to analyze layer number, defects, doping and strain of graphene samples. At the same time, it is an intriguing playground to study the optical properties of graphene. [less ▲] Detailed reference viewed: 313 (5 UL) Raman spectroscopy of graphite intercalation compounds: Charge transfer, strain, and electron–phonon coupling in graphene layers; Wirtz, Ludger ; in Physica Status Solidi B. Basic Research (2014), 251(12), 23372355 Graphite intercalation compounds (GICs) are an interesting and highly studied field since 1970’s. It has gained renewed interest since the discovery of superconductivity at high temperature for CaC6 and ... [more ▼] Graphite intercalation compounds (GICs) are an interesting and highly studied field since 1970’s. It has gained renewed interest since the discovery of superconductivity at high temperature for CaC6 and the rise of graphene. Intercalation is a technique used to introduce atoms or molecules into the structure of a host material. Intercalation of alkali metals in graphite has shown to be a controllable procedure recently used as a scalable technique for bulk production of graphene, and nano-ribbons by induced exfoliation of graphite. It also creates supra-molecular interactions between the host and the intercalant, inducing changes in the electronic, mechanical, and physical properties of the host. GICs are the mother system of intercalation also employed in fullerenes, carbon nanotubes, graphene, and carbon-composites. We will show how a combination of Raman and ab-initio calculations of the density and the electronic band structure in GICs can serve as a tool to elucidate the electronic structure, electron–phonon coupling, charge transfer, and lattice parameters of GICs and the graphene layers within. This knowledge of GICs is of high importance to understand superconductivity and to set the basis for applications with GICs, graphene and other nano-carbon based materials like nanocomposites in batteries and nanoelectronic devices. [less ▲] Detailed reference viewed: 222 (11 UL) Raman spectroscopy of rare-earth orthoferrites RFeO3 (R=La, Sm, Eu, Gd Tb, Dy); Guennou, Mael ; et alin PHYSICAL REVIEW B (2016), 94(21), We report a Raman scattering study of six rare-earth orthoferrites La, Sm, Eu, Gd, Tb, Dy. The use of extensive polarized Raman scattering of SmFeO3 and first-principles calculations enable the assignment ... [more ▼] We report a Raman scattering study of six rare-earth orthoferrites La, Sm, Eu, Gd, Tb, Dy. The use of extensive polarized Raman scattering of SmFeO3 and first-principles calculations enable the assignment of the observed phonon modes to vibrational symmetries and atomic displacements. The assignment of the spectra and their comparison throughout the whole series allow correlating the phonon modes with the orthorhombic structural distortions of RFeO3 perovskites. In particular the positions of two specific A(g) modes scale linearly with the two FeO6 octahedra tilt angles, allowing the distortion to be tracked throughout the series. At variance with literature, we find that the two octahedra tilt angles scale differently with the vibration frequencies of their respective A(g) modes. This behavior, as well as the general relations between the tilt angles, the frequencies of the associated modes, and the ionic radii are rationalized in a simple Landau model. The reported Raman spectra and associated phonon-mode assignment provide reference data for structural investigations of the whole series of orthoferrites. [less ▲] Detailed reference viewed: 99 (2 UL)![]() Raman spectroscopy of single-wall BN nanotubes; Wirtz, Ludger ; et alin AIP Conference Proceedings (2003), (685), 384-388 We present results on the vibrational properties of BN-SWNTs together with a study of the synthesis material by transmission electron microscopy. Phonon modes have been investigated by Raman spectroscopy ... [more ▼] We present results on the vibrational properties of BN-SWNTs together with a study of the synthesis material by transmission electron microscopy. Phonon modes have been investigated by Raman spectroscopy with laser excitation wavelengths in the range from 363.8 to 676.44 nm. The assignment of the modes is guided by ab-initio calculations. [less ▲] Detailed reference viewed: 92 (0 UL) Raman spectroscopy of single-wall boron nitride nanotubes; ; et al in Nano Letters (2006), 6(8), 1812-1816 Single-wall boron nitride nanotubes samples synthesized by laser vaporization of a hexagonal BN target under a nitrogen atmosphere are studied by UV and visible Raman spectroscopy. We show that resonant ... [more ▼] Single-wall boron nitride nanotubes samples synthesized by laser vaporization of a hexagonal BN target under a nitrogen atmosphere are studied by UV and visible Raman spectroscopy. We show that resonant conditions are necessary for investigating phonon modes of BNNTs. Raman excitation in the UV (229 nm) provides preresonant conditions, allowing the identification of the A(1) tangential mode at 1370 cm(-1). This is 5 cm(-1) higher than the E-2g mode in bulk h-BN. Ab initio calculations show that the lower frequency of bulk h-BN with respect to large diameter nanotubes and the single sheet of h-BN is related to a softening of the sp2 bonds in the bulk due to interlayer interaction. [less ▲] Detailed reference viewed: 111 (2 UL) Raman spectroscopy on mechanically exfoliated pristine graphene ribbons; Reichardt, Sven ; et alin Physica Status Solidi B. Basic Research (2014), 251(12), 2551--2555 Detailed reference viewed: 128 (3 UL)![]() Raman spectroscopy on single- and few-layer graphene; ; et al in AIP Conference Proceedings (2007), 893 We report on Raman measurements of single- and few-layer graphene flakes. Raman mapping in combination with scanning force microscopy allows us to locally relate the thickness of the graphite flake with ... [more ▼] We report on Raman measurements of single- and few-layer graphene flakes. Raman mapping in combination with scanning force microscopy allows us to locally relate the thickness of the graphite flake with the spectral properties. It turns out that the width of the D' line is highly sensitive to the transition from single- to double-layer graphene. The defect-induced D line is found to be most prominent at steps between sections of different height and along the edge of the graphite flake. [less ▲] Detailed reference viewed: 120 (0 UL) |
||