Éditeur(s) : HAL CCSD IEEE
Résumé : International audience
In this paper, we present an evolutionary approach for extracting a model of flood prediction from hydrological data observed timely on water heights in a river watershed. Since this kind of data recorded by sensors on river basins is highly scarce and hopefully much unbalanced between cases of floods and non-floods, we have adopted the notion of aggregate variables which values are computed as aggregates on raw data. An evolutionary algorithm is involved to allow selecting the best sets - juries of classifiers- of such variables as predictive variables. Two real hydrological data sets are trained and they both show the efficiency of the method compared to traditional solutions for prediction.
IEEE Congress on Evolutionary Computation 2012
Brisbane, Australia

hal-00840739
https://hal.archives-ouvertes.fr/hal-00840739
DOI : 10.1109/CEC.2012.6252897

Résumé : Cette thèse concerne la modélisation des dynamiques de maladies foliaires de cultures pérennes tropicales à différentes échelles spatiales. Cette approche de modélisation est appliquée à la cercosporiose noire du bananier. Il s’agit d’explorer et de déterminer les paramètres environnementaux et de résistance de l’hôte ayant une influence significative sur la dynamique spatiotemporelle de la maladie et d’apporter des éléments associés au contrôle de la cercosporiose noire.Deux modèles ont été développés dans le cadre de cette thèse. La dynamique épidémiologique au niveau de la plante est décrite par un modèle mécaniste décomposé en un modèle de croissance de la plante et un module épidémiologique décrivant le cycle épidémique du pathogène. L’architecture du bananier est prise en compte par le biais de compartiments foliaires positionnés dans l’espace. Le modèle a été validé par un jeu de données indépendant. Les expérimentations numériques et l’analyse de sensibilité du modèle réalisées par les méthodes de Morris et de e-FAST ont permis de mieux comprendre le fonctionnement épidémique de cette maladie et d’identifier les paramètres influant le plus la dynamique épidémique en particulier la vitesse d’extension des lésions, la durée d’incubation et l’efficacité d’infection .L’approche bayésienne a permis de prendre en compte l’information a priori disponible pour ces trois paramètres sur lesquels a porté l’inférence statistique. L’analyse de sensibilité du modèle a également permis d’identifier l’influence de deux paramètres liés à la croissance de la plante : le nombre de feuilles sur une plante et le rythme d’émission foliaire Un modèle de dynamique spatio-temporelle de la cercosporiose noire a été développé à l’échelle d’un territoire à partir d’enquêtes effectuées enMartinique pendant la période d’invasion de cette maladie de septembre 2010 à mai 2012. Les données récoltées étant censurées, l’inférence des paramètres du modèle a été réalisée dans un cadre bayésien en utilisant un algorithme d’augmentation de données. Le modèle et l’inférence développés permettent de reconstruire la dynamique spatio-temporelle de l’invasion et de prédire la fin d’invasion sur le territoire.Les deux modèles de dynamiques spatio-temporelles développés à deux échelles spatiales différentes ont permis d’acquérir des informations importantes pour construire des outils de conception de méthode de contrôle de la cercosporiose noire des bananiers.
This thesis concerns the modeling of the dynamics of foliar diseases of tropical tree crops at different spatial scales. This modeling approach is applied to black Sigatoka of banana. It is to explore and determine the environmental parameters and host resistance has a significant influence on the spatiotemporal dynamics of the disease and to provide elements associated with the control of Sigatoka noire.Deux models were developed as part of this thesis. The epidemiological dynamics at the plant is described by a mechanistic model decomposed into a growth model of the plant and an epidemiological unit describing the pathogen epidemic cycle. The architecture of the banana is taken into account through foliar compartments positioned in space. The model was validated by an independent data set. Numerical experiments and model sensitivity analysis performed by the methods of Morris and e-FAST enabled to better understand the functioning of this epidemic disease and identify the parameters affecting the most dynamic epidemic especially speed extension of the lesions, the incubation period and efficiency of Bayesian .L'approche infection allowed to take into account prior information available for these three parameters that were the statistical inference. The model sensitivity analysis also identified the influence of two parameters related to plant growth: the number of leaves on a plant and rate of leaf emission A dynamic model of space-time Black Sigatoka has been developed at the scale of a territory from surveys enMartinique during the period of invasion of the disease from September 2010 to May 2012. The data collected being censored, inference of model parameters has was performed in a Bayesian framework, using a data augmentation algorithm. The model developed and inference possible to reconstruct the spatiotemporal dynamics of the invasion and predict the end of the invasion territoire.Les two spatio-temporal dynamics models developed at two different spatial scales has been gained important information for build tools design method of control of black Sigatoka of banana.

http://www.theses.fr/2015AGUY0835/document

Éditeur(s) : HAL CCSD
Résumé : International audience
This work proposes an original approach using mutual information and Pareto curve jointly for feature selection. Mutual information is used to estimate dependency criterion between features and classes and redundancy criterion between features taken two by two. Unlike some studies, these criteria are used simultaneously to compute Pareto curve and determine the optimal feature sets. This approach is tested on more reference data. Several clustering algorithms are used to compute classification accuracy. The obtained results show the importance of our tools and its ability to select the best feature sets that give the better description.
GRETSI Proceedings
Bordeaux, France

hal-00602260
https://hal.archives-ouvertes.fr/hal-00602260

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

Éditeur(s) : HAL CCSD
Résumé : International audience
The method we propose is a new approach to the problem of extbf{satellite constellation design}. The main difficulties of this field are the size of the solution space, the computation time of the optimization criterion and the lack of information to analyse and improve a solution. Our model bypasses some of these obstacles by using an extbf{inverse approach} where services to be fulfilled are highlighted. The algorithm is composed of different layers solving different problems. We first use the specifications of the services to compensate for lack of information returned by the criterion to analyse and improve a solution. An orbit database introduces an expert knowledge linked to a robust estimator to set consistent orbital parameters. We use a partial evaluation of the constellation to reduce the criterion computation time. The highest level of the algorithm is a metaheuristic layer that drives the search across the wide solution space with a probabilistic tabu search.
ICAI Proceeding

hal-00634752
https://hal.archives-ouvertes.fr/hal-00634752

Éditeur(s) : HAL CCSD EDP Sciences
Résumé : We propose and we justify a Monte-Carlo algorithm which solves a spatially homogeneous kinetic equation of Boltzmann type that models the fusion reaction between a deuterium ion and a tritium ion, and giving an α particle and a neutron. The proposed algorithm is validated with the use of explicit solutions of the kinetic model obtained by replacing the fusion cross-section by a Maxwellian cross section.
ISSN: 1270-900X

hal-00689821
https://hal.archives-ouvertes.fr/hal-00689821
https://hal.archives-ouvertes.fr/hal-00689821/document
https://hal.archives-ouvertes.fr/hal-00689821/file/proc_versionHAL.pdf

Éditeur(s) : Centre français du littoral
Résumé : This study presents the first results of an ongoing research program that aims to derive thematic maps of the intertidal lagoon of Arcachon (SW Atlantic coast of France) from space imagery. At this stage, we present the results of the tests performed on Formosat-2 imagery. Several classification strategies have been tried with the aim to discriminate the sedimentary facies and the main species of the lagoon. The classes calculated by an optimized unsupervised clustering method, based on simulated annealing global minimization technique, were used to train a supervised classification algorithm. This approach allowed us to derive a bio-sedimentary map which overall accuracy is about 84%. Also, this method has the advantage of describing the fraction of vegetation cover of the seagrass.
Cette étude présente les premiers résultats d’un vaste programme de recherche qui ambitionne de définir une méthode de cartographie des platiers intertidaux du Bassin d’Arcachon (France) par télédétection. Les résultats présentés ici concernent les tests effectués sur l’imagerie Formosat-2 dans le but de discriminer les sédiments et les principales espèces végétales de la lagune. Le résultat d'une classification non supervisée optimisée par la méthode du recuit simulé est ensuite utilisé pour entraîner une classification par minimum de distance. Cette approche permet d’obtenir une carte des faciès bio-sédimentaires dont la précision est estimée à plus de 84 %. Cette méthode présente, en outre, un excellent potentiel pour la détermination du pourcentage de couvert végétal des herbiers à zostères.
Paralia (1760-8716) (Centre français du littoral), 2008 , N. 10 , P. 563-572

Droits : info:eu-repo/semantics/openAccess
http://archimer.ifremer.fr/doc/00254/36572/35110.pdf
DOI:10.5150/jngcgc.2008.054-L
http://archimer.ifremer.fr/doc/00254/36572/

Éditeur(s) : HAL CCSD Public Library of Science
Résumé : International audience
Various types of genome and gene similarity networks along with their characteristics have been increasingly used for retracing different kinds of evolutionary and ecological relationships. Here, we present a new polynomial time algorithm and the corresponding software (BRIDES) to provide characterization of different types of paths existing in evolving (or augmented) similarity networks under the constraint that such paths contain at least one node that was not present in the original network. These different paths are denoted as Breakthroughs , Roadblocks, Impasses, Detours, Equal paths, and Shortcuts. The analysis of their distribution can allow discriminating among different evolutionary hypotheses concerning genomes or genes at hand. Our approach is based on an original application of the popular shortest path Dijkstra's and Yen's algorithms.
ISSN: 1932-6203

Droits : http://creativecommons.org/licenses/by/
hal-01361753
http://hal.upmc.fr/hal-01361753
http://hal.upmc.fr/hal-01361753/document
http://hal.upmc.fr/hal-01361753/file/journal.pone.0161474.PDF
DOI : 10.1371/journal.pone.0161474

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Two splitting procedures for solving equilibrium problems involving the sum of two bifunctions are proposed and their convergence is established under mild assumptions.
Journal of Mathematical Analysis and applications

hal-00779982
https://hal.univ-antilles.fr/hal-00779982
DOI : 10.1016/j.jmaa.2009.06.005

Éditeur(s) : HAL CCSD Wiley
Résumé : International audience
The brown algal genus Padina (Dictyotales, Phaeophyceae) is distributed worldwide in tropical and temperate seas. Global species diversity and distribution ranges, however, remain largely unknown. Species-level diversity was reassessed using DNA-based, algorithmic species delineation techniques based on cox3 and rbcL sequence data from 221 specimens collected worldwide. This resulted in estimates ranging from 39 to 61 putative species (ESUs), depending on the technique as well as the locus. We discuss the merits, potential pitfalls, and evolutionary and biogeographic significance of algorithmic species delineation. We unveil patterns whereby ESUs are in all but one case restricted to either the Atlantic or Indo-Pacific Ocean. Within ocean basins we find evidence for the vast majority of ESUs to be confined to a single marine realm. Exceptions, whereby ESUs span up to three realms, are located in the Indo-Pacific Ocean. Patterns of range-restricted species likely arise by repeated founder events and subsequent peripatric speciation, hypothesized to dominate speciation mechanisms for coastal marine organisms in the Indo-Pacific. Using a three-gene (cox3, psaA and rbcL), relaxed molecular clock phylogenetic analysis we estimated divergence times, providing a historical framework to interpret biogeographic patterns.
ISSN: 0022-3646

hal-01545323
https://hal.archives-ouvertes.fr/hal-01545323
DOI : 10.1111/jpy.12027

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
This article is devoted to the presentation of a multilevel method using finite differences that is well adapted for solving Stokes and Navier-Stokes problems in primitive variables. We use Uzawa type algorithms to solve the saddle point problems arising from spatial discretization by staggered grids and a semi-explicit temporal scheme. By means of a new change of basis operator, the two-dimensional velocity and pressure fields of an M.A.C mesh are gathered in a hierarchical order, into several grids preserving the M.A.C property on each of them. These new hierarchical unknowns, called Staggered Incremental Unknowns (SIU), allow us to develop techniques which reduce the cost of the resolution of either Stokes or generalized Stokes problems. An experimental estimation of the condition number of the inner matrix is given, and justifies the preconditioning effect of the staggered incremental unknowns.
ISSN: 0168-9274

hal-00770268
https://hal.univ-antilles.fr/hal-00770268
DOI : 10.1016/S0168-9274(99)00044-6

Éditeur(s) : HAL CCSD
Résumé : Session "Articles"
National audience
Dans cet article, nous présentons une méthode robuste et précise pour détecter 17 points caractéristiques du visage sur des images expressives. Une nouvelle architecture multi-résolution basée sur les récents algorithmes multi-noyau est introduite. Les patches de faibles résolutions codent les informations globales du visage donnant lieu à une détection grossière mais robuste du point désiré. Les patches de grandes résolutions quant à eux utilisent les détails locaux afin d'affiner cette localisation. En combinant une détection indépendante de points et des informations a priori sur les distributions de points, nous proposons un détecteur robuste aux changements d'expressions ainsi qu'aux variations d'illuminations. Ce système a été évalué sur plusieurs bases de données de la littérature. Les résultats présentés améliorent les performances des détecteurs de l'état de l'art
Actes de la conférence RFIA 2012
Lyon, France

hal-00656522
https://hal.archives-ouvertes.fr/hal-00656522
https://hal.archives-ouvertes.fr/hal-00656522/document
https://hal.archives-ouvertes.fr/hal-00656522/file/rfia2012_submission_99.pdf

Éditeur(s) : HAL CCSD
Résumé : The convex feasibility problem (CFP) is to find a member of the intersection of finitely many closed convex sets in Euclidean space. When the intersection is empty, one can minimize a proximity function to obtain an approximate solution to the problem. The split feasibility problem (SFP) and the split equality problem (SEP) are generalizations of the CFP. The approximate SFP (ASFP) and approximate SEP (ASEP) involve finding only approximate solutions to the SFP and SEP, respectively. We present here the SSEA, a simultaneous iterative algorithm for solving the ASEP. When this algorithm is applied to the ASFP it resembles closely, but is not equivalent to, the CQ algorithm. The SSEA involves orthogonal projection onto the given closed convex sets. The relaxed SSEA (RSSEA) is an easily implementable variant of the SSEA that uses orthogonal projection onto half-spaces at each step to solve the SEP. The perturbed version of the SSEA (PSSEA) is similar to the RSSEA, but uses orthogonal projection onto a sequence of epi-convergent closed convex sets.
https://hal.univ-antilles.fr/hal-00776640

Droits : info:eu-repo/semantics/OpenAccess
hal-00776640
https://hal.univ-antilles.fr/hal-00776640
https://hal.univ-antilles.fr/hal-00776640/document
https://hal.univ-antilles.fr/hal-00776640/file/CBAM10.pdf

Éditeur(s) : HAL CCSD
Résumé : 17 pages
We look at the number of solutions of an equation of the form f_1*f_2*...*f_k=a in a finite field, where each f_i is a multilinear polynomial. We use two methods to construct a solution of this problem for the cases a=0, a<>0, and we generally get a semi-explicit formula. We show that this formula can generate a more efficient algorithm than the traditional algorithm which consists to make a systematic computation. We also give explicit formulas in some special cases, and an application of our main result to the search of the weight hierarchy of the multilinear code with separated variables.
https://hal.univ-antilles.fr/hal-00772478

hal-00772478
https://hal.univ-antilles.fr/hal-00772478
ARXIV : math.NT/0603057

Éditeur(s) : HAL CCSD Lavoisier
Résumé : International audience
In this paper, we present an heuristic based approach for feature selection in the context of flood prediction. Features are complex variables that represent aggregate values. We apply a preprocessing method on data in order to elicit relevant information that could not be easily accessible initially because it is split through several lines of a dataset. A genetic algorithm is used in order to search for the features that may prove the best performances for flood prediction.
ISSN: 1633-1311

hal-00840738
https://hal.archives-ouvertes.fr/hal-00840738
DOI : 10.3166/isi.16.3.113-139

Éditeur(s) : HAL CCSD
Résumé : The research presented in this thesis aims at determining the possibility of partial or total automatic ship detection in High Resolution optical satellite images. At first, we expose the industrial context of our study, which fixes the constraints in terms of performances and processing time. Then, we present the available satellite data for our study (SPOT 5 panchromatic images of 5m per pixel resolution). Then we detail our approach for ship detection : the first stage is a pre-detection of targets that gives us candidates. The second stage is a precise segmentation of each candidate and the third stage is a classification of candidates. Our pre-detection method is based on the Bayesian decision theory. This pre-detection supplies a set of regions of interests (ROI). Among them, we have ships, but also false alarms due to clouds and to crests of waves. The purpose of the following stage is to eliminate the false alarms and to characterize the pre-detected ships. Then, we detail the method used to segment candidates. It uses active contours and also uses the Bayesian decision theory. After a brief description of the characteristics extracted for the classication of ships, we present the classifier that we used. We complete the ship characterization by presenting two methods of wave ship detection. The first approach uses active contours and the second is based on an a contrario approach. We compare the results of these two approaches. To complete the study and replace it in the industrial and operational context, we present experiments of the complete algorithm on complete images to estimate the performances and the processing time of the whole method.
Les travaux présentés dans cette thèse ont pour but de déterminer la possibilité d'automatiser, partiellement ou totalement, la détection de bateaux à partir d'images satellitaires optiques haute résolution. Nous exposons tout d'abord le contexte industriel dans lequelle s'insère notre étude, qui fixe les contraintes en termes de performances et de temps de traitement. Puis, nous présentons les données satellitaires disponibles dans le cadre de notre étude (images SPOT 5 panchromatique de résolution 5m). Après avoir illustré le contenu des scènes, nous présentons la démarche adoptée pour la détection des navires : une phase de pré-détection suivie d'une étape de segmentation et d'une étape de classification. La méthode de pré-détection utilisée s'appuie sur la théorie bayésienne de la décision. Après avoir présenté le modèle classique, nous discutons du modèle que nous avons utilisé. Cette pré-détection fournit un ensemble de régions d'intérêts. Parmi elles, nous n'avons pas uniquement des bateaux, nous avons également des fausses alarmes dues aux nuages et aux crêtes de vagues. Le but de l'étape suivante est à la fois d'éliminer les fausses alarmes et de caractériser les bateaux pré-détectés. Ensuite, nous détaillons la méthode utilisée pour segmenter les bateaux. Elle s'appuie sur des contours actifs et utilise également la théorie bayésienne de la décision. Après avoir décrit les caractéristiques extraites pour la classification des bateaux, nous présentons le classifieur utilisé. Nous complétons la caractérisation des navires en présentant deux méthodes de détection des sillages différentes des méthodes classiques. La première approche s'appuie sur l'utilisation de contours actifs et la seconde est basée sur une approche a contrario. Nous comparons les résultats des ces deux approches. An de compléter l'étude et se replacer dans le contexte industriel et opérationnel de la thèse, nous présentons des expérimentations de l'ensemble de l'algorithme sur des images complètes pour évaluer les performances et le temps de traitement.
https://tel.archives-ouvertes.fr/tel-00560119

tel-00560119
https://tel.archives-ouvertes.fr/tel-00560119
https://tel.archives-ouvertes.fr/tel-00560119/document
https://tel.archives-ouvertes.fr/tel-00560119/file/These_NadiaProia_Version_Finale.pdf

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Based on the very recent work by Censor and Segal (2009), and inspired by Xu (2006), Zhao and Yang (2005) , and Bauschke and Combettes (2001), we introduce and analyze an algorithm for solving the split common fixed-point problem for the wide class of quasi-nonexpansive operators in Hilbert spaces. Our results improve and develop previously discussed feasibility problems and related algorithms.
ISSN: 0362-546X

hal-00776645
https://hal.univ-antilles.fr/hal-00776645
DOI : 10.1016/j.na.2011.03.041

Éditeur(s) : HAL CCSD Springer
Résumé : International audience
Numerous methods have been proposed in order to perform clustering from social networks. While significant works have been carried out on the design of new approaches, able to search for various kinds of clusters, a major challenge concerns the scalability of these approaches. Indeed, given the mass of data that can now be collected from online social networks, particularly from social platforms, it is important to have efficient methods for exploring and analyzing these very large amount of data. One of the recent social network clustering approaches is the extraction of conceptual links, a new approach that performs link clustering by exploiting both the structure of the network and attributes of nodes to identify strong links between groups of nodes in which nodes share common attributes. In this paper, we focus on the optimization of the search for conceptual links. In particular, we propose PALM, a parallel algorithm that aims to improve the efficiency of the extraction by simultaneously exploring several areas of the search space. For this purpose, we begin by demonstrating that the solution space forms a concept lattice. Then, we propose an approach that explores in parallel the branches of the lattice while reducing the search space based on various properties of conceptual links. We demonstrate the efficiency of the algorithm by comparing the performances with the original extraction approach. The results obtained show a significant gain on the computation time.
28th International Conference on Database and Expert Systems Applications (DEXA)
Lyon, France

hal-01637554
https://hal.archives-ouvertes.fr/hal-01637554
DOI : 10.1007/978-3-319-64471-4_21

Éditeur(s) : HAL CCSD Oxford University Press (OUP)
Résumé : This paper presents an inversion method using the interseismic velocity field to determine effective rigidity of the lithosphere. The method is based on the minimization of a cost function defined as the quadratic measure of the difference between measured and modelled velocity fields on a discrete set of points. The continuous mapping of the rigidity is fulfilled with a limited set of parameters and the forward solution is achieved using a plane stress finite element code. The computation of the cost function gradient in the parameters' space allows one to iteratively find the best parameters set through a suitable optimization algorithm.;We first design a benchmark including an abrupt rigidity variation that cannot be described by a continuous function. For such a case, we show that increasing the number of parameters is a way to accurately describe sharp variations of the rigidity map. Then, we use a dense GPS velocity field over the southwestern United States to estimate the corresponding rigidity variations for different spatial resolutions of the parameters' grid. We analyse the conceptual and practical difficulties associated with our methodology. Finally, rigidity maps obtained by our inversion method in southwestern United States and particularly across the San Andreas Fault System are reviewed and compared to current plate rigidity estimates and geophysical data over this area.
ISSN: 0956-540X

hal-00669897
https://hal.archives-ouvertes.fr/hal-00669897
DOI : 10.1111/j.1365-246X.2011.05192.x