Éditeur(s) : HAL CCSD
Résumé : International audience
Some morphological and biochemical taxonomic markers are also affecting factors of dynamic mechanical properties of wood. Thus, could these properties reflect the classification/phylogeny of taxa? This work is a first insight into this question. It relied on the gathering (through experimental campaigns and exhaustive literature review) of a large database on the viscoelastic (i.e. including damping) vibrational properties of 445 woody species. relationship between damping was confirmed at a wide interspecific scale, but described no more than 40% of not in a way that could be easily related to the phylogenetic tree. Damping was a bit more -Papilionoideae and, to a lesser extent, Lauraceae, Cupressaceae and Moraceae) damping than average, independently While for some other families no clear characteristics could be observed at least with the present number of represented species. In the future, increasing the amount of data and compiling anatomical and chemical markers / affecting factors will allow further analysis at sub-family levels, and a better understanding of this wide topic.
6th Plant Biomechanics Conference PBM09
Cayenne, French Guiana

hal-00808398
https://hal.archives-ouvertes.fr/hal-00808398
https://hal.archives-ouvertes.fr/hal-00808398/document
https://hal.archives-ouvertes.fr/hal-00808398/file/Mechanical_damping_wood_Bremaud_al.pdf

Éditeur(s) : Elsevier
Résumé : In order to provide reliable scientific advice and support for fisheries management, it is necessary to evaluate the biological and economic sustainability of complex fisheries, such as multi-species multi-fleet fisheries. Existing policy-screening modelling tools are not fully suitable in this purpose due to either an over-simplified description of population dynamics, or due to the lack of consideration of economic aspects. In this paper, we present a package that enables quantitative bioeconomic assessment of management scenarios. Population dynamics is described through spatially- and seasonally-explicit models. Exploitation dynamics is characterized by several fishing activities with specific spatial and seasonal features, and practiced by several kinds of vessels with specific technical characteristics. Exploitation costs and revenues are considered at several levels: the fishing trip, the fishing unit (vessel and crew), and the vessel owner. The model is generic and can be used for different types of fisheries. A database is attached to the software for the storage and updating of information for each fishery. This includes the specification of model dimensions and of the parameters describing populations and exploitation. Several model assumptions regarding either population or exploitation may be adapted to suit a specific fishery. Both policies and corresponding fishers' response may be interactively specified through JAVA (TM) scripts. This version of ISIS-Fish allows for the calculation of biological and economic consequences of a range of policies, including conventional ones like catch and effort controls, and alternative policies such as marine protected areas. To facilitate policy-screening in a high-dimension parameter space, the software includes features, like interfaces for sensitivity analysis and simulation queues. (C) 2009 Elsevier B.V. All rights reserved.
Ecological Modelling (0304-3800) (Elsevier), 2009-04 , Vol. 220 , N. 7 , P. 1013-1033

Droits : 2009 Elsevier B.V. All rights reserved.
http://archimer.ifremer.fr/doc/2009/publication-6782.pdf
DOI:10.1016/j.ecolmodel.2009.01.007
http://archimer.ifremer.fr/doc/00000/6782/

Éditeur(s) : Littoral 2004, Aberdeen, Scotland, UK, September 2004
Résumé : NAMIBIE (a french acronym for Integrated Multimedia Navigator Within Coastal Environment Data Bases) is a specific tool which first concept has been established by Ifremer and has been developped by an association between research in the field of coastal environment, innovation in the field of new information and communication technologies and services in the field of coastal integrated management and impact studies. The objectives of this tool are to facilitate research, direct access and display of relevant multimedia coastal integrated data and information according to an interactive, widely opened and playfull interface. The principles are based upon the fact that sustainable decision making needs an optimized unterstanding of coastal processes and better consultation between actors. In this purpose the development of the tool has considered that the end-user should and could be free to look after the data and the information he needs for a particular site or environment of interest and for a specific ICZM question according to easy navigation interfaces. That is why NAMIBIE concept is based upon : - a coherent 3D representation of the coasts (i.e. from catchement areas to continental shelf and deep sea), which considers a coherent DEM from land to sea and textural representation of the environment using satellite imagery textural displays, - a 3D navigator which allows the user to fly or dive over and within the coastal environment of interest, - a multimedia database (text, graphs, image, maps, geographic objects, modelling results...) specific to the territory of interest, - an interactive interface which helps in finding relevant information associated to specific questions. The technologies used include VRML data format, integration of standard GIS format (shp, grid, img) within the navigator, integration and display of modelling data. The production chain is realized in C++ and Java scripts allow dialogue between web pages and 3D windows. Applications are presented on the coasts of Languedoc Roussillon region (French Mediterranean border) in the framework of ICZM problematics

Droits : info:eu-repo/semantics/openAccess
http://archimer.ifremer.fr/doc/00089/20055/17696.pdf
http://archimer.ifremer.fr/doc/00089/20055/

Éditeur(s) : HAL CCSD
Résumé : National audience
Lorsque le volume des données est trop important pour qu'elles soient stockées dans une base de données, ou lorsque leur taux d'arrivée est rapide, les Systèmes de Gestion de Flux de Données (SGFD) permettent de capturer des flux d'enregistrements structurés et de les interroger à la volée par des requêtes permanentes (exécutées de façon continue). Mais les SGFD ne conservent pas l'historique des flux qui est perdu à jamais. Cette communica- tion propose une définition formelle de ce que devrait être un résumé généraliste de flux de données. La notion de résumé généraliste est liée à la capacité de répondre à des requêtes variées et de réaliser des tâches variées de fouille de données, en utilisant le résumé à la place du flux d'origine. Une revue de plusieurs approches de résumés est ensuite réalisée dans le cadre de cette définition. Enfin, l'intégration de telles approches de résumés dans un SGFD est discutée, à la fois en termes de construction et d'utilisation.
EGC: Extraction et Gestion des Connaissances
Hammamet, Tunisia

hal-00502017
https://hal.archives-ouvertes.fr/hal-00502017
https://hal.archives-ouvertes.fr/hal-00502017/document
https://hal.archives-ouvertes.fr/hal-00502017/file/resume_generaliste.pdf

Éditeur(s) : HAL CCSD Oxford University Press (OUP)
Résumé : International audience
OBJECTIVES: To compare virological effectiveness in patients who continued on a virologically successful first-line boosted protease inhibitor (PI)-containing combination antiretroviral therapy (cART) regimen or who switched to a PI-free cART including efavirenz, nevirapine or abacavir. METHODS: From the French Hospital Database on HIV, we selected 439 patients with undetectable viral load (VL) on a first-line boosted PI-containing cART regimen who switched to a PI-free combination including efavirenz, nevirapine or abacavir. Each of these patients was matched with three patients who continued to take their first-line cART regimen, on the basis of gender, age, CD4 cell count, VL, date of cART initiation and the duration of VL undetectability. Time to virological failure (VF) was analysed with Kaplan-Meier curves and Cox models. RESULTS: The 12 month probabilities of VF were 3.7% and 5.7% in non-switch and switch patients, respectively, and 3.9%, 7.2% and 9.0% in patients switching to efavirenz-, nevirapine- and abacavir-containing cART, respectively. After adjustment, only patients switching to abacavir-containing cART had a higher risk of VF than non-switch patients (adjusted hazard ratio, 1.99; 95% confidence interval, 1.05-3.79). CONCLUSIONS: Switching from a virologically successful first-line boosted PI-containing cART regimen to a non-nucleoside reverse transcriptase inhibitor-containing cART regimen containing either efavirenz or nevirapine is virologically safe, while switching to abacavir-containing cART should be avoided.
ISSN: 0305-7453

hal-00875284
https://hal.archives-ouvertes.fr/hal-00875284
DOI : 10.1093/jac/dkr208
PUBMED : 21636583

Éditeur(s) : HAL CCSD
Résumé : Half-way between optimization and astronautic, this research study deals with satellites constel- lations design problems. In order to find the required number of satellites and to correctly set their position, such are the technical challenges of this thesis. To minimise the cost and reduce the time such are the economical challenges the space industries are confronted in a daily manner. The major difficulties of these problems are the size and the characteristics of the search space The irregularity of the criterions The mathematical and physical heterogeneity of parameters forbids the use of classical algorithms The evaluation of a solution, which uses a time consuming evalua- tion without returning pertinent information about the good or bad properties of the constellations, forbid a massive exploration of the search space. From this constatation and several preliminary studies a new approach is born. Based on a better use of the simulation and on a simplification of the criterions, the algorithm is composed of several levels and uses different optimization techniques it integrates a knowledge database on the orbits and a numerical optimization process both orchestrated by a metaheuristic algorithm. This new approach tries to bypass the main drawbacks of the field with a decomposition of the problem.
A mi-chemin entre l optimisation et l astronautique ce travail de recherche s attaque au problème du design de constellations de satellites Déterminer le nombre de satellites nécessaires pour ré- pondre à un besoin et les positionner de manière correcte tels sont les enjeux techniques de cette thèse Minimiser les coûts et les délais tels sont les enjeux économiques auxquels est confronté le domaine spatial au quotidien Les difficultés majeures de ce problème sont d une part la taille et les caractéristiques de l espace de recherche; D autre part la non-régularité des critères utilisés et la nature mathématique et phy- sique hétérogène de l ensemble des paramètres (paramètres entiers et paramètres réels) interdisant l utilisation d algorithmes classiques; Enfin l évaluation des solutions qui met en oeuvre une simu- lation coûteuse en temps et en calculs sans retourner d information pertinente sur les bonnes ou mauvaises propriétés de la constellation évaluée interdit une exploration massive de l espace de recherche De ces constatations et de diverses études préliminaires est née une nouvelle approche Basé sur une meilleure exploitation de la simulation et sur une simplification des critères utilisés l algorithme composé de plusieurs niveaux met en oeuvre plusieurs techniques d optimisation En intégrant une base de connaissance sur les orbites et une méthode numérique de recherche d optimum le tout or- chestré par un algorithme métaheuristique cette nouvelle approche tente de résoudre les difficultés intrinsèques au domaine d application en décomposant le problème.
https://tel.archives-ouvertes.fr/tel-00634769

tel-00634769
https://tel.archives-ouvertes.fr/tel-00634769
https://tel.archives-ouvertes.fr/tel-00634769/document
https://tel.archives-ouvertes.fr/tel-00634769/file/grandchamp_phd_2001.pdf

Éditeur(s) : The 11th Pacific Science Inter-Congress : Pacific Countries and their Ocean: Facing Local and Global Changes. March 2 - 6, 2009 Tahiti, French Polynesia
Résumé : New Caledonian shrimp farming feasibility studies started in the mid-seventies. Today ca. 2000 tons of Litopenaeus stylirostris, introduced in 1978, are semi-intensively produced each year. Since fifteen years, production has been impacted by two seasonal vibriosis, the “Summer” syndrome caused by Vibrio nigripulchritudo in relation to pond bottom and water column quality, and the “Winter Syndrome” caused by Vibrio penaeicida in relation to drastic temperature drops . With the ambition of elucidating processes possibly leading to mortality outbreaks in ponds, Ifremer has been conducting a research program since 2003 under the DESANS project (2003-2006) and the present DEDUCTION project, equally funded by the South and North Provinces and by the Government of New Caledonia. This program has been designed in a multidisciplinary approach integrating the suspected multifactorial origin of shrimp mortalities. Environment quality in ponds (either sediment or water column) plays a major role in the shrimp ecophysiological status, influencing sensitivity to pathogens . DEDUCTION focuses on a better knowledge of phytoplankton composition and biogeochemical fluxes at the sediment-water interface in ponds. Impact of farm discharges is also investigated. In addition, a more fundamental work is conducted to understand the pathogenicity mechanisms of these two Vibrio species but also to prevent the occurrence of viral infections through a RNA interference approach. Markers of the ecophysiological status of broodstock, larvae, post-larvae and reared animals, such as enzymes testifying from an oxidative stress, are developed. Gain of performance (resistance to pathogens and growth) were obtained on hybrids of the New Caledonian and introduced Hawaiian strains. Finally the database Stylog is exploited in view of discriminating “good” from “bad” rearings on the basis of relevant indicators.
Les études de faisabilité de l'aquaculture de crevettes ont débuté en Nouvelle-Calédonie dans les années soixante dix. Aujourd'hui environ 2000 tonnes de Litopenaeus stylirostris, introduite en 1978, sont produites chaque année de façon semi-intensive. Depuis quinze ans la production a été impactée par deux vibrioses : le syndrome d'été provoqué par Vibrio nigripulchritudo en relation avec la qualité de la tranche d'eau et des fonds de bassins et le syndrome d'hiver provoqué par Vibrio penaeicida en relation avec des chutes rapides de température. Dans le but d'élucider les processus entrainant des mortalités dans les élevages, Ifremer a conduit un programme de recherche depuis 2003 dans le cadre du projet DESANS (2003-2006) puis du présent projet DEDUCTION co-financé par les provinces Nord et Sud et le Gouvernement de la Nouvelle-Calédonie. Ce programme se construit dans une démarche multidisciplinaire intégrant les origines supposées multifactorielles des mortalités de crevettes. La qualité environnementale dans les bassins (les fonds et la colonne d'eau) joue un rôle majeur dans l'état physiologique des animaux, influençant leur sensibilité aux pathogènes. DEDUCTION se focalise sur une meilleure compréhension de la composition phytoplanctonique et des flux bio-géochimiques à l'interface eau/sédiment dans les bassins.L'impact des rejets des fermes est également étudié. En outre, un travail plus fondamental est conduit pour connaître les mécanismes de pathogénicité de ces deux espèces de Vibrio mais aussi pour prévenir les occurences d'infections virales par des approches ARN Interférence. Des marqueurs de l'état physiologique des oeufs, larves, post-larves et animaux en élevage, tels que les enzymes du stress oxydant sont utilisés.Les gains de performance (résistance aux pathogènes et croissance) sont obtenus sur des hybrides issus du croisement de la souche locale avec une souche hawaienne. Finalement la base de données STYLOG est exploitée dans le but de distinguer les bonnes des mauvaises pratiques sur la base d'indicateurs fiables.

Droits : info:eu-repo/semantics/openAccess
http://archimer.ifremer.fr/doc/00143/25438/23591.pdf
http://archimer.ifremer.fr/doc/00143/25438/

Éditeur(s) : HAL CCSD
Résumé : National audience
Nous présentons les premières phases d'un travail orienté vers la construction d'un thésaurus trilingue dédié à la biodiversité en méditerranée. Notre démarche se veut résolument synthétique et exploite à posteriori les dires d'experts accessibles depuis les publications scientifiques faisant référence dans le domaine. Les termes clés, qui sont à la base du socle conceptuel nécessaire à une meilleure compréhension de la biodiversité en méditerranée, sont acquis au travers d'une approche rigoureuse de fouille de texte qui est détaillée ici. Les premiers résultats obtenus sont rapportés et évalués. Les experts du domaine seront ensuite sollicités pour aider à articuler les termes au travers de structures hiérarchiques et ainsi retranscrire leur savoir sous la forme d'un thésaurus dédié à la biodiversité dans le bassin méditerranéen.
INFORSID
Montpellier, France

lirmm-00723586
http://hal-lirmm.ccsd.cnrs.fr/lirmm-00723586
http://hal-lirmm.ccsd.cnrs.fr/lirmm-00723586/document

Résumé : This paper deals with a small-scale fisheries sample survey conducted in 1987 in Martinique. In 25 main sites, direct observations were made of numbers of fishing trips and of their characteristics fishing effort, weight and composition of catch, ... In 106 secondary sites, only numbers of trips were indirectly recorded, through informal interviews of fishermen or local inhabitants. The structure of the computer database is described in the appendix (data files, programs). NOT CONTROLLED OCR
Ce document décrit le système d'enquêtes utilisé en 1987 dans les sites de débarquement de la pêche artisanale martiniquaise. Des observations directes sont réalisées dans 25 sites principaux, portant sur le nombre et les caractéristiques des sorties de pêche: effort, poids et composition des prises, ... Dans 106 sites secondaires, seuls les nombres de sorties sont mesurés, par enquête indirecte auprès des pêcheurs ou des riverains. La structure de la base de données informatique (fichiers de données et programmes) est décrite en annexe. OCR NON CONTRÔLE

Droits : info:eu-repo/semantics/openAccess
http://archimer.ifremer.fr/doc/1988/rapport-3884.pdf
http://archimer.ifremer.fr/doc/00000/3884/

Éditeur(s) : HAL CCSD Oxford University Press (OUP)
Résumé : International audience
OBJECTIVES: To better select for patients who most likely will benefit from cardiac surgery among those with infective endocarditis (IE), we aimed to identify preoperative markers associated with poor outcome after cardiac surgery for IE, and to evaluate the accuracy of European System for Cardiac Operative Risk Evaluation (EuroSCORE) II to predict mortality. METHODS: We enrolled all adult patients who underwent cardiac surgery during the acute phase of definite IE (Duke Criteria) in two referral centres for cardiac surgery. Patients were identified through intensive care unit (ICU) electronic databases, and data were collected from medical charts on standardized questionnaire. RESULTS: Between 2002 and 2013, 149 patients (117 males), with a median age of 64 years [interquartile range 52-73], fulfilled the inclusion criteria. Main complications before surgery were left ventricular dysfunction (23%), central nervous system symptomatic events (34%) and septic shock (24%). Most patients (95%) presented with valve regurgitation, and 49% had perivalvular abscess. Surgery was performed with a median delay of 12 days [5-24] after IE diagnosis, and mean EuroSCORE II was 15.8 (13.4-18.1). In-hospital mortality was 21%. Preoperative variables associated with mortality in multivariate analysis were obesity [odds ratio (OR) 3.67 [1.10-12.19], P = 0.03], vegetation extgreater15 mm (OR 6.72 [1.46-30.98], P = 0.01), septic shock (OR 4.87 [1.67-14.28], P = 0.004) and mechanical prosthetic valve IE (OR 4.99 [1.72-28.57], P = 0.007). EuroSCORE II underestimated mortality in patients with predicted mortality over 10%. CONCLUSION: Factors independently predictive of mortality after cardiac surgery for IE are obesity, septic shock, large vegetation and a mechanical prosthetic valve IE. EuroSCORE II underestimates post-cardiac surgery mortality in patients with IE
ISSN: 1010-7940

hal-01169802
https://hal-univ-rennes1.archives-ouvertes.fr/hal-01169802
DOI : 10.1093/ejcts/ezv223
PUBMED : 26116921

Éditeur(s) : HAL CCSD American Chemical Society
Résumé : Three new monoterpene indole alkaloids (13) have been isolated from the bark of Geissospermum laeve-, together with the known alkaloids (-)-leuconolam (4), geissolosimine (5), and geissospermine (6). The structures of 1-3 were elucidated by analysis of their HRMS and NMR spectroscopic data. The absolute configuration of geissolaevine (I). was deduced from the comparison of experimental and theoretically calculated ECD spectra. The isolation workflow was guided by a molecular networking-based dereplication strategy using an in-house database of monoterpene indole alkaloids. In addition, five known compounds previously undescribed in the Geissospermum genus were dereplicated from the G. laeve alkaloid extract network and were assigned with various levels of identification confidence. The antiparasitic activities against Plasmodium falciparum and Leishmania donovani as well as the Ortotoxic activity against the MRC-5 cell line were determined for compounds 1-5.
ISSN: 0163-3864

hal-01605065
https://hal.archives-ouvertes.fr/hal-01605065
DOI : 10.1021/acs.jnatprod.6b01013
PRODINRA : 402605
PUBMED : 28282127

Éditeur(s) : HAL CCSD Frontiers Media
Résumé : International audience
We present high resolution observations of microstructure and texture evolution during dynamic recrystallization (DRX) of ice polycrystals deformed in the laboratory at high temperature (≈0.98Tm). Ice possesses a significant viscoplastic anisotropy that induces strong strain heterogeneities, which result in an early occurrence of DRX mechanisms. It is therefore a model material to explore these mechanisms. High resolution c-axis measurements at sample scale by optical techniques and full crystallographic orientation measurements by cryo-Electron Back Scattering Diffraction (EBSD) provide a solid database for analyzing the relative impact of the macroscopic imposed stress vs. the local and internal stress fields on DRX mechanisms. Analysis of misorientation gradients in the EBSD data highlights a heterogeneous dislocation distribution, which is quantified by the Nye tensor estimation. Joint analyses of the dislocation density maps and microstructural observations highlight spatial correlation between high dislocation density sites and the onset of nucleation taking place by grain-boundary bulging, subgrain rotation or by the formation of kink-bands.
ISSN: 2296-6463

hal-01515284
https://hal.univ-lorraine.fr/hal-01515284
DOI : 10.3389/feart.2015.00081

Éditeur(s) : HAL CCSD Oxford University Press (OUP): Policy B1 - Oxford Open Option B Oxford University Press (OUP)
Résumé : International audience
BACKGROUNDA very large proportion of the literature on the endocrine disruptors categorized as anti-androgens deals with phthalates, which are produced in large amounts for use as plastic emollients and additives. In this review, we bring together and analyse work on the effects of phthalates in animals and humans at different stages of their development to assess whether or not their possible anti-androgenic properties represent a significant threat to human health.METHODSThe database PubMed was systematically searched for all English language articles until July 2013 in each subject area discussed.RESULTSWe provide an up-to-date exhaustive, comparative and critical assessment of both in vivo and in vitro studies undertaken to explore the effects of phthalates on the human testis from fetal life to adulthood. These results are compared and discussed in the light of the key data reported in the literature for mice and rats.CONCLUSIONSThe current literature highlights the fact that (i) there is a huge difference between the number of studies performed in animals and in humans, with many fewer for humans; (ii) there are differences in the way rats, mice, primates and humans respond to phthalates, for reasons that need to be further explored; (iii) more work is required to clarify the contradictions, in the few existing human epidemiological studies at all stages of development, which may be partly explained by varying methods of exposure assessment; (iv) in accordance with recent findings in rodents, it cannot be excluded that transgenerational effects of phthalates and/or epigenetic changes exist in humans; (v) a number of methodological limitations need to be solved for the in vitro and xenografting models using human fetal testis to fulfil their 'missing link' role between epidemiological studies in humans and rodent models; and (vi) epidemiological and in vitro studies generally converge sufficiently to conclude that phthalate anti-androgenicity is plausible in adult men.
ISSN: 1355-4786

inserm-00869540
http://www.hal.inserm.fr/inserm-00869540
DOI : 10.1093/humupd/dmt050
PUBMED : 24077978

Éditeur(s) : HAL CCSD
Résumé : International audience
The tropical rainforest of Amazonia is one of the most species-rich ecosystems on earth, with an estimated 16000 tree species. Due to this high diversity, botanical identification of trees in the Amazon is difficult, even to genus, often requiring the assistance of parataxonomists or taxonomic specialists. Advances in informatics tools offer a promising opportunity to develop user-friendly electronic keys to improve Amazonian tree identification.Here, we introduce an original multi-access electronic key for the identification of 389 tree genera occurring in French Guiana terra-firme forests, based on a set of 79 morphological characters related to vegetative, floral and fruit characters. Its purpose is to help Amazonian tree identification and to support the dissemination of botanical knowledge to non-specialists, including forest workers, students and researchers from other scientific disciplines.The electronic key is accessible with the free access software Xper², and the database is publicly available on figshare: https://figshare.com/s/75d890b7d707e0ffc9bf
ISSN: 1314-2011

hal-01392306
https://hal-sde.archives-ouvertes.fr/hal-01392306
DOI : 10.3897/phytokeys.68.8707

Éditeur(s) : HAL CCSD
Résumé : LPSC14037
Scientific data collected with modern sensors or dedicated detectors exceed very often the perimeter of the initial scientific design. These data are obtained more and more frequently with large material and human efforts. A large class of scientific experiments are in fact unique because of their large scale, with very small chances to be repeated and to superseded by new experiments in the same domain: for instance high energy physics and astrophysics experiments involve multi-annual developments and a simple duplication of efforts in order to reproduce old data is simply not affordable. Other scientific experiments are in fact unique by nature: earth science, medical sciences etc. since the collected data is "time-stamped" and thereby non-reproducible by new experiments or observations. In addition, scientific data collection increased dramatically in the recent years, participating to the so-called "data deluge" and inviting for common reflection in the context of "big data" investigations. The new knowledge obtained using these data should be preserved long term such that the access and the re-use are made possible and lead to an enhancement of the initial investment. Data observatories, based on open access policies and coupled with multi-disciplinary techniques for indexing and mining may lead to truly new paradigms in science. It is therefore of outmost importance to pursue a coherent and vigorous approach to preserve the scientific data at long term. The preservation remains nevertheless a challenge due to the complexity of the data structure, the fragility of the custom-made software environments as well as the lack of rigorous approaches in workflows and algorithms. To address this challenge, the PREDON project has been initiated in France in 2012 within the MASTODONS program: a Big Data scientific challenge, initiated and supported by the Interdisciplinary Mission of the National Centre for Scientific Research (CNRS). PREDON is a study group formed by researchers from different disciplines and institutes. Several meetings and workshops lead to a rich exchange in ideas, paradigms and methods. The present document includes contributions of the participants to the PREDON Study Group, as well as invited papers, related to the scientific case, methodology and technology. This document should be read as a "facts finding" resource pointing to a concrete and significant scientific interest for long term research data preservation, as well as to cutting edge methods and technologies to achieve this goal. A sustained, coherent and long term action in the area of scientific data preservation would be highly beneficial.
http://hal.in2p3.fr/in2p3-00959072

in2p3-00959072
http://hal.in2p3.fr/in2p3-00959072
http://hal.in2p3.fr/in2p3-00959072/document
http://hal.in2p3.fr/in2p3-00959072/file/PREDON_Scientific_Data_Preservation_2014.pdf

Éditeur(s) : HAL CCSD Institute of Electrical and Electronics Engineers
Résumé : International audience
Tailoring nearest neighbors algorithms to boosting is an important problem. Recent papers study an approach, UNN, which provably minimizes particular convex surrogates under weak assumptions. However, numerical issues make it necessary to experimentally tweak parts of the UNN algorithm, at the possible expense of the algorithm's convergence and performance. In this paper, we propose a lightweight alternative algorithm optimizing proper scoring rules from a very broad set, and establish formal convergence rates under the boosting framework that surprisingly compete with those known for UNN. It is an adaptive Newton-Raphson algorithm, which belongs to the same lineage as the popular Gentle Adaboost. To the best of our knowledge, no such boosting-compliant convergence rates were previously known for these algorithms. We provide experiments on a dozen domains, including the challenging Caltech and SUN computer vision databases. They display that GNNB significantly outperforms UNN, both in terms of convergence rate and quality of the solution obtained, and GNNB provides a simple and efficient contender to techniques that can be used on very large domains, like stochastic gradient descent -- for which little is known to date. Experiments include a divide-and-conquer improvement of GNNB which exploits the link with proper scoring rules optimization.
ISSN: 0162-8828

hal-00958809
https://hal.archives-ouvertes.fr/hal-00958809
https://hal.archives-ouvertes.fr/hal-00958809/document
https://hal.archives-ouvertes.fr/hal-00958809/file/PAMI.pdf
DOI : 10.1109/TPAMI.2014.2307877

Éditeur(s) : HAL CCSD
Résumé : International audience
Objective: In order to determine how risk factors of death among HIV-infected patients in French Guiana interact with one another and to define high risk population segments, an alternative statistical method to Cox proportional hazards models was used. Patients and methods: Data from HIV+ patients followed in the three main hospitals of French Guiana were used. Data were extracted from the prospective French Hospital Database for HIV (FHDH). To examine the nature of the relationships between time of death since the first consultation and a set of predictor variables (age at enrolment, gender, nationality, CD4 count at enrolment, presence or absence of HAART at enrolment, a declared addiction, and mode of acquisition of the virus) a survival tree analysis was used. Survival rates at 5 years and 10 years and their 95% confidence interval were also calculated. Results: This analysis permitted to reveal a new specific subgroup of HIV+ patients particularly at risk in French Guiana not already mentioned in the literature: the HIV+ female patients with a declared addiction. Conclusions: Prevention and care in this vulnerable segment of the HIV+ population identified by this methodology should be reinforced.
J AIDS Clinic Res

Droits : info:eu-repo/semantics/OpenAccess
inserm-00914956
http://www.hal.inserm.fr/inserm-00914956
http://www.hal.inserm.fr/inserm-00914956/document
http://www.hal.inserm.fr/inserm-00914956/file/2012_Hanf_-_Hierarchysur Determinants_Underlying_Death_among_HIV-Infected_patients_in_French_Guiana.pdf
DOI : 10.4172/2155-6113.1000172

Éditeur(s) : Mercator Ocean
Résumé : Face to the lack of data to assess precisely the spatial distribution of catches and fishing effort and for the environmental characterization of the fishing area, Ifremer has been implemented since 2005 a new project, Recopesca. It consists in fitting out a sample of voluntary fishing vessels with sensors recording data on fishing effort (and at mid-terms catches) and physical parameters such as temperature or salinity. Recopesca aims at setting up a network of sensors, for scientific purposes, to collect data and improve resources assessment and diagnostics on fisheries, and environmental data required for ecosystem-based management initiatives. The challenge was to develop sensors with no trouble for the fishermen, tough enough to be fixed up on fishing gears, self powered and autonomous. Insofar as the sample of targeted vessels intends to be representative of all the metiers and fleets, the sensors are modular and scalable to collect new data. Different sensors have been implemented: (i) a temperature-salinity sensor, able to record physical parameters, depth and duration of immersion, for passive and active gears, and (ii) a specific sensor to record number or length of passive gears. A GPS monitors the position of the vessels and the temperature or salinity profiles and series. Each sensor is equipped with a radio device transferring the data to a receiver on-board, called “concentrator” that sends the data to Ifremer central databases by GPRS. An anti-rolling weigh-scale has been developed and is currently on test to record catches per species and fishing operation. The presentation will show the first data and results of this participative approach.
Mercator Ocean - Quarterly Newsletter (Mercator Ocean), 2010-04 , N. 37 , P. 40-48

Droits : 2010 Mercator Ocean
http://archimer.ifremer.fr/doc/00024/13500/10514.pdf
http://archimer.ifremer.fr/doc/00024/13500/

Éditeur(s) : HAL CCSD
Résumé : International audience
It is an admitted fact that mainstream boosting algorithms like AdaBoost do not perform well to estimate class conditional probabilities. In this paper, we analyze, in the light of this problem, a recent algorithm, unn, which leverages nearest neighbors while minimizing a convex loss. Our contribution is threefold. First, we show that there exists a subclass of surrogate losses, elsewhere called balanced, whose minimization brings simple and statistically efficient estimators for Bayes posteriors. Second, we show explicit convergence rates towards these estimators for unn, for any such surrogate loss, under a Weak Learning Assumption which parallels that of classical boosting results. Third and last, we provide experiments and comparisons on synthetic and real datasets, including the challenging SUN computer vision database. Results clearly display that boosting nearest neighbors may provide highly accurate estimators, sometimes more than a hundred times more accurate than those of other contenders like support vector machines.
European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD)
Bristol, United Kingdom

hal-00702771
https://hal.archives-ouvertes.fr/hal-00702771
https://hal.archives-ouvertes.fr/hal-00702771/document
https://hal.archives-ouvertes.fr/hal-00702771/file/ecml12-dnbnb-sub.pdf

Éditeur(s) : HAL CCSD Springer
Résumé : Third International Workshop, MLMI 2012, Held in Conjunction with MICCAI 2012
International audience
Universal Nearest Neighbours (unn) is a classifier recently proposed, which can also effectively estimates the posterior probability of each classification act. This algorithm, intrinsically binary, requires the use of a decomposition method to cope with multiclass problems, thus reducing their complexity in less complex binary subtasks. Then, a reconstruction rule provides the final classification. In this paper we show that the application of unn algorithm in conjunction with a reconstruction rule based on the posterior probabilities provides a classification scheme robust among different biomedical image datasets. To this aim, we compare unn performance with those achieved by Support Vector Machine with two different kernels and by a k Nearest Neighbours classifier, and applying two different reconstruction rules for each of the aforementioned classification paradigms. The results on one private and five public biomedical datasets show satisfactory performance.
Machine Learning in Medical Imaging
Nice, France

hal-00958909
https://hal.archives-ouvertes.fr/hal-00958909
https://hal.archives-ouvertes.fr/hal-00958909/document
https://hal.archives-ouvertes.fr/hal-00958909/file/Miccai.pdf
DOI : 10.1007/978-3-642-35428-1_15