Papers by Christian Farinetto
<p>CHG: Chlorhexidine Alcohol, PVI: Povidone Alcohol, AE: Adverse event, CRBSI: Catheter-re... more <p>CHG: Chlorhexidine Alcohol, PVI: Povidone Alcohol, AE: Adverse event, CRBSI: Catheter-related bloodstream infection, CT: Catheter, G+S: Semipermeable transparent dressing.</p
<p>Costs per patient per Markov state (input parameters).</p
<p>Non-homogeneous Semi-Markov Chain simulation from observed data (CLEAN)–simulated global... more <p>Non-homogeneous Semi-Markov Chain simulation from observed data (CLEAN)–simulated global patient—ICU-time horizon: 100 days.</p
<p>Resources (catheters, dressings, antiseptic solutions) per patient—Statistical unit: The... more <p>Resources (catheters, dressings, antiseptic solutions) per patient—Statistical unit: The global patient with catheterization (alive, discharge or dead).</p
Communications in Statistics - Theory and Methods, 2010
We consider the problem of parameter estimation for an inhomogeneous Poisson process observed on ... more We consider the problem of parameter estimation for an inhomogeneous Poisson process observed on the time interval [0, τ]. We introduce the minimum L1-norm estimator of the unknown parameter and study the asymptotical behaviors of the estimates when the number of observations increases. It is established that this estimator is consistent and we show that the corresponding differences converge to certain variables. These limit variables are asymptotically normal as τ tends to infinity.
Statistical Inference for Stochastic Processes
<p>Cost-effectiveness results from observed data (CLEAN database)–Observed global patient—I... more <p>Cost-effectiveness results from observed data (CLEAN database)–Observed global patient—ICU-Time Horizon: 100 days.</p
Japanese Journal of Statistics and Data Science
Annals of the Institute of Statistical Mathematics
We present a detection problem where several spatially distributed sensors observe Poisson signal... more We present a detection problem where several spatially distributed sensors observe Poisson signals emitted from a single source of unknown position. The measurements at each sensor are modeled by independent inhomogeneous Poisson processes. A method based on Bayesian change-point estimation is proposed to identify the location of the source's coordinates. The asymptotic behavior of the Bayesian estimator is studied. In particular the consistency and the asymptotic efficiency of the estimator are analyzed. The limit distribution and the convergence of the moments are also described. The similar statistical model could be used in GPS localization problems.
Journal de gestion et d'économie médicales, 2011
Objectif : L’objectif de cette etude est d’evaluer sous un angle economique l’impact d’une struct... more Objectif : L’objectif de cette etude est d’evaluer sous un angle economique l’impact d’une structure d’accompagnement et d’education a la sante des assures sociaux de la Sarthe. Cette structure, denommee « Club Sante Active » (CSA), a ete creee par la Caisse Primaire d’Assurance Maladie de la Sarthe en janvier 2006. L’analyse economique est proposee dans la perspective de l’Assurance maladie. Il s’agit d’une etude observationnelle retrospective fondee sur l’exploitation des donnees de remboursement de soins sollicites entre 2005 et 2009 par les assures sociaux du regime general de la Sarthe. Methode : 563 patients ont ete inclus dans le groupe CSA et 832 dans le groupe temoin. Une methode statistique comparable a celle du score de propension a ete utilisee afin d’assurer la comparabilite des deux groupes etudies a partir de la prise en compte de neuf covariables disponibles. Aussi, une analyse statistique des series temporelles de consommation medicale des assures de chaque groupe a-t-elle ete realisee afin d’evaluer l’impact economique tendanciel du Club Sante Active. Les depenses de sante remboursees par l’Assurance maladie (couts directs medicaux et non medicaux, indemnites journalieres) constituent le critere final de jugement de l’efficacite economique de la strategie evaluee et la nonadhesion au Club Sante Active definit la strategie de reference. Resultats : Une moindre evolution tendancielle des consommations par rapport au groupe temoin (hors CSA) a ete relevee chez les membres du CSA sur les categories de personnes et groupes de prestations suivants : les personnes en affection de longue duree (ALD) âgees de 40 a 60 ans pour les analgesiques et le recours aux soins de medecine generale, les personnes en ALD âgees de plus de 60 ans pour les analgesiques et les inhibiteurs de la pompe a proton (IPP), les personnes sans ALD âgees de 40 a 60 ans pour les IPP, les psychotropes, le recours aux soins de kinesitherapie et le nombre d’indemnites journalieres versees. A l’inverse, un seul cas d’evolution superieure pour les adherents du Club a ete recense : il s’agit du recours aux soins infirmiers pour les personnes en ALD de plus de 60 ans. Ce dernier resultat doit toutefois etre mis en perspective avec la particularite de la demographie medicale des infirmiers en Sarthe, ou l’acces a ce type de soins n’est a priori pas satisfaisant, le departement etant l’un des moins bien pourvus de toute la France. Conclusion : Les signes d’evolution constates sur les soins de ville ne sont pas la contrepartie d’un report de depenses sur les soins hospitaliers, ces prestations ayant ete controlees. Ces resultats tendent ainsi a confirmer que l’accompagnement des assures peut aboutir a une modification significative des comportements de recours au systeme de soins, notamment sur le champ des soins de ville.
We propose a new method for the economic evaluation of hybrid fixed wireless/optical access netwo... more We propose a new method for the economic evaluation of hybrid fixed wireless/optical access networks. These networks are costeffective solutions in countries with a limited infrastructure. The evaluation of the cost function of such networks requires a detailed description of their spatial features that is usually not available and unsuitable for economic analysis. Our model captures the essential spatial characteristics of optical ring networks by representing its components as a family of random object generated by a spatial Poisson process. We give bounds on the cost of feeder rings and the exact cost function of distribution subnetworks as functional of the intensities of Poisson processes. 1
We propose a new method for the economic evaluation of hybrid xed wireless/optical access network... more We propose a new method for the economic evaluation of hybrid xed wireless/optical access networks. These networks are coste ective solutions in countries with a limited infrastructure. The evaluation of the cost function of such networks requires a detailed description of their spatial features that is usually not available and unsuitable for economic analysis. Our model captures the essential spatial characteristics of optical ring networks by representing its components as a family of random object generated by a spatial Poisson process. We give bounds on the cost of feeder rings and the exact cost function of distribution subnetworks as functional of the intensities of Poisson processes.
Communications in Statistics - Theory and Methods
ABSTRACT We consider the problem of testing whether realizations of a spatial Poisson process com... more ABSTRACT We consider the problem of testing whether realizations of a spatial Poisson process come from a circular intensity model or from an elliptical intensity model. We describe the behavior of the Generalized Likelihood Ratio and Wald tests, in the asymptotics of large samples. The power functions are studied under local alternatives and results of numerical simulations comparing them to the Neyman–Pearson envelope are presented.
Communications in Statistics - Theory and Methods
ABSTRACT Consider an inhomogeneous Poisson process X on [0, T] whose unk-nown intensity function ... more ABSTRACT Consider an inhomogeneous Poisson process X on [0, T] whose unk-nown intensity function “switches” from a lower function g* to an upper function h* at some unknown point ϑ* that has to be identified. We consider two known continuous functions g and h such that g*(t) ⩽ g(t) < h(t) ⩽ h*(t) for 0 ⩽ t ⩽ T. We describe the behavior of the generalized likelihood ratio and Wald’s tests constructed on the basis of a misspecified model in the asymptotics of large samples. The power functions are studied under local alternatives and compared numerically with help of simulations. We also show the following robustness result: the Type I error rate is preserved even though a misspecified model is used to construct tests.
We propose a new method for the economic evaluation of hybrid fixed wireless/optical access netwo... more We propose a new method for the economic evaluation of hybrid fixed wireless/optical access networks. These networks are costeffective solutions in countries with a limited infrastructure. The evaluation of the cost function of such networks requires a detailed description of their spatial features that is usually not available and unsuitable for economic analysis. Our model captures the essential spatial characteristics of optical ring networks by representing its components as a family of random object generated by a spatial Poisson process. We give bounds on the cost of feeder rings and the exact cost function of distribution subnetworks as functional of the intensities of Poisson processes.
Annales De L Isup, 2002
On considere le probleme d'estimation du parametre de l'intensite d'un processus de P... more On considere le probleme d'estimation du parametre de l'intensite d'un processus de Poisson spatial inhomogene par la methode de la distance minimale. On suppose que l'intensite est discontinue le long d'une courbe dependant d'un parametre inconnu. Les resultats etablis montrent la consistance et la normalite asymptotique de l'estimateur de distance minimale.
C R Acad Sci Ser I Math, 2001
Uploads
Papers by Christian Farinetto