Université de Montréal

About: Université de Montréal is a education organization based out in Montreal, Quebec, Canada. It is known for research contribution in the topics: Population & Poison control. The organization has 45641 authors who have published 100476 publications receiving 4004007 citations. The organization is also known as: University of Montreal & UdeM.

An Improved Spectroscopic Analysis of DA White Dwarfs from the Sloan Digital Sky Survey Data Release 4

Abstract: We present an improved spectroscopic and photometric analysis of hydrogen-line DA white dwarfs from the Sloan Digital Sky Survey Data Release 4 (SDSS DR4) based on model atmospheres that include improved Stark broadening profiles with non-ideal gas effects. We also perform a careful visual inspection of all spectroscopic fits with high signal-to-noise ratios (S/Ns > 12) and present improved atmospheric parameters (T eff and log g) for each white dwarf. Through a comparison of spectroscopic and photometric temperatures, we report the discovery of 35 DA+DB/DC double degenerate candidates and two helium-rich DA stars. We also determine that a cutoff at S/N = 15 optimizes the size and quality of the sample for computing the mean mass of DA white dwarfs, for which we report a value of 0.613 M ☉. We compare our results to previous analyses of the SDSS DR4 and find a good agreement if we account for the shift produced by the improved Stark profiles. Finally, the properties of DA white dwarfs in the SDSS are weighed against those of the Villanova White Dwarf Catalog sample of Gianninas et al. We find systematically lower masses (by about 3% on average), a difference that we trace back to the data reduction procedure of the SDSS. We conclude that a better understanding of these differences will be important to determine the absolute temperature scale and mean mass of DA white dwarfs.

On Variational Bounds of Mutual Information

Abstract: Estimating and optimizing Mutual Information (MI) is core to many problems in machine learning; however, bounding MI in high dimensions is challenging. To establish tractable and scalable objectives, recent work has turned to variational bounds parameterized by neural networks, but the relationships and tradeoffs between these bounds remains unclear. In this work, we unify these recent developments in a single framework. We find that the existing variational lower bounds degrade when the MI is large, exhibiting either high bias or high variance. To address this problem, we introduce a continuum of lower bounds that encompasses previous bounds and flexibly trades off bias and variance. On high-dimensional, controlled problems, we empirically characterize the bias and variance of the bounds and their gradients and demonstrate the effectiveness of our new bounds for estimation and representation learning.

Deep Learning of Representations: Looking Forward

Abstract: Deep learning research aims at discovering learning algorithms that discover multiple levels of distributed representations, with higher levels representing more abstract concepts. Although the study of deep learning has already led to impressive theoretical results, learning algorithms and breakthrough experiments, several challenges lie ahead. This paper proposes to examine some of these challenges, centering on the questions of scaling deep learning algorithms to much larger models and datasets, reducing optimization difficulties due to ill-conditioning or local minima, designing more efficient and powerful inference and sampling procedures, and learning to disentangle the factors of variation underlying the observed data. It also proposes a few forward-looking research directions aimed at overcoming these challenges.

Prevention of Atrial Fibrillation Report From a National Heart, Lung, and Blood Institute Workshop

Abstract: The National Heart, Lung, and Blood Institute convened an expert panel April 28 to 29, 2008, to identify gaps and recommend research strategies to prevent atrial fibrillation (AF). The panel reviewed the existing basic scientific, epidemiological, and clinical literature about AF and identified opportunities to advance AF prevention research. After discussion, the panel proposed the following recommendations: (1) enhance understanding of the epidemiology of AF in the population by systematically and longitudinally investigating symptomatic and asymptomatic AF in cohort studies; (2) improve detection of AF by evaluating the ability of existing and emerging methods and technologies to detect AF; (3) improve noninvasive modalities for identifying key components of cardiovascular remodeling that promote AF, including genetic, fibrotic, autonomic, structural, and electrical remodeling markers; (4) develop additional animal models reflective of the pathophysiology of human AF; (5) conduct secondary analyses of already-completed clinical trials to enhance knowledge of potentially effective methods to prevent AF and routinely include AF as an outcome in ongoing and future cardiovascular studies; and (6) conduct clinical studies focused on secondary prevention of AF recurrence, which would inform future primary prevention investigations.