Bayesian inference and model comparison for random choice structures

We complete the development of a testing ground for axioms of discrete stochastic choice. Our contribution here is to develop new posterior simulation methods for Bayesian inference, suitable for a class of prior distributions introduced by McCausland and Marley (2013). These prior distributions are joint distributions over various choice distributions over choice sets of di fferent sizes. Since choice distributions over di fferent choice sets can be mutually dependent, previous methods relying on conjugate prior distributions do not apply. We demonstrate by analyzing data from a previously reported experiment and report evidence for and against various axioms.

Azabicycloalkanone synthesis by transannular cyclization

Une stratégie de synthèse efficace de différents composés de type azabicyclo[X.Y.0]alkanone fonctionnalisés a été développée. La stratégie synthétique implique la préparation de dipeptides par couplage avec des motifs vinyl-, allyl-, homoallyl- et homohomoallylglycine suivi d’une réaction de fermeture de cycle par métathèse permettant d’obtenir des lactames macrocycliques de 8, 9 et 10 membres, qui subissent une iodolactamisation transannulaire menant à l’obtention de mimes peptidiques bicycliques portant un groupement iode. Des couplages croisés catalysés par des métaux de transition ont été développés pour la synthèse d’acides aminés ω-insaturés énantiomériquement purs à partir de l’iodoanaline. L’étude du mécanisme suggère que l’iodure subit une attaque du coté le moins stériquement encombré de la lactame macrocyclique insaturée pour mener à l’obtention d’un intermédiaire iodonium. La cyclisation se produit ensuite par une route minimisant les interactions diaxiales et la tension allylique. L’iodolactamisation des différentes lactames macrocycliques insaturées a mené à l’obtention regio- et diastéréosélective d’acides aminés 5,5- et 6,6-iodobicycicliques. De plus, une imidate azabicyclo[4.3.1]alkane pontée de type anti-Bredt fut synthétisée à partir d’une lactame macrocyclique insaturé à neuf membres. Les analyses cristallographiques et spectroscopiques des macrocycles à 8, 9 et 10 membres, du composé iodobicyclique 5,5 ainsi que de l’imidate pontée, montrent bien le potentiel de ces dipeptides rigidifiés de servir en tant que mimes des résidus centraux de tours β de type I, II’, II et VI.

Simulation-Based Finite-and Large-sample Inference Methods in Multivariate Regressions and Seemingly Unrelated Regressions

In the context of multivariate regression (MLR) and seemingly unrelated regressions (SURE) models, it is well known that commonly employed asymptotic test criteria are seriously biased towards overrejection. in this paper, we propose finite-and large-sample likelihood-based test procedures for possibly non-linear hypotheses on the coefficients of MLR and SURE systems.

Spectrum simulation of rough and nanostructured targets from their 2D and 3D image by Monte Carlo methods

Corteo is a program that implements Monte Carlo (MC) method to simulate ion beam analysis (IBA) spectra of several techniques by following the ions trajectory until a sufficiently large fraction of them reach the detector to generate a spectrum. Hence, it fully accounts for effects such as multiple scattering (MS). Here, a version of Corteo is presented where the target can be a 2D or 3D image. This image can be derived from micrographs where the different compounds are identified, therefore bringing extra information into the solution of an IBA spectrum, and potentially significantly constraining the solution. The image intrinsically includes many details such as the actual surface or interfacial roughness, or actual nanostructures shape and distribution. This can for example lead to the unambiguous identification of structures stoichiometry in a layer, or at least to better constraints on their composition. Because MC computes in details the trajectory of the ions, it simulates accurately many of its aspects such as ions coming back into the target after leaving it (re-entry), as well as going through a variety of nanostructures shapes and orientations. We show how, for example, as the ions angle of incidence becomes shallower than the inclination distribution of a rough surface, this process tends to make the effective roughness smaller in a comparable 1D simulation (i.e. narrower thickness distribution in a comparable slab simulation). Also, in ordered nanostructures, target re-entry can lead to replications of a peak in a spectrum. In addition, bitmap description of the target can be used to simulate depth profiles such as those resulting from ion implantation, diffusion, and intermixing. Other improvements to Corteo include the possibility to interpolate the cross-section in angle-energy tables, and the generation of energy-depth maps.