Browse ORBi by ORBi project

- Background
- Content
- Benefits and challenges
- Legal aspects
- Functions and services
- Team
- Help and tutorials

Comparison between principal component analysis and independent component analysis in EEG modelling ; Lambert, Philippe in Biometrical Journal = Biometrische Zeitschrift (2007), 49 Principal Component Analysis (PCA) is a classical technique in statistical data analysis, feature extraction and data reduction, aiming at explaining observed signals as a linear combination of orthogonal ... [more ▼] Principal Component Analysis (PCA) is a classical technique in statistical data analysis, feature extraction and data reduction, aiming at explaining observed signals as a linear combination of orthogonal principal components. Independent Component Analysis (ICA) is a technique of array processing and data analysis, aiming at recovering unobserved signals or ‘sources’ from observed mixtures, exploiting only the assumption of mutual independence between the signals. The separation of the sources by ICA has great potential in applications such as the separation of sound signals (like voices mixed in simultaneous multiple records, for example), in telecommunication or in the treatment of medical signals. However, ICA is not yet often used by statisticians. In this paper, we shall present ICA in a statistical framework and compare this method with PCA for electroencephalograms (EEG) analysis.We shall see that ICA provides a more useful data representation than PCA, for instance, for the representation of a particular characteristic of the EEG named event-related potential (ERP). [less ▲] Detailed reference viewed: 36 (1 ULg)Robust specification of the roughness penalty prior distribution in spatially adaptive Bayesian P-splines models ; Lambert, Philippe in Computational Statistics & Data Analysis (2007), 51 The potential important role of the prior distribution of the roughness penalty parameter in the resulting smoothness of Bayesian Psplines models is considered. The recommended specification for that ... [more ▼] The potential important role of the prior distribution of the roughness penalty parameter in the resulting smoothness of Bayesian Psplines models is considered. The recommended specification for that distribution yields models that can lack flexibility in specific circumstances. In such instances, these are shown to correspond to a frequentist P-splines model with a predefined and severe roughness penalty parameter, an obviously undesirable feature. It is shown that the specification of a hyperprior distribution for one parameter of that prior distribution provides the desired flexibility. Alternatively, a mixture prior can also be used. An extension of these two models by enabling adaptive penalties is provided. The posterior of all the proposed models can be quickly explored using the convenient Gibbs sampler. [less ▲] Detailed reference viewed: 30 (3 ULg)Functional ANOVA with random functional effects: an application to event-related potentials modelling for electroencephalograms analysis ; Lambert, Philippe in Statistics in Medicine (2006), 25 The di erential e ects of basic visual or auditory stimuli on electroencephalograms (EEG), named event related potentials (ERPs), are often used to evaluate the impact of treatments on brain performances ... [more ▼] The di erential e ects of basic visual or auditory stimuli on electroencephalograms (EEG), named event related potentials (ERPs), are often used to evaluate the impact of treatments on brain performances. In the present paper, we propose a P-splines based model that can be used to evaluate treatment e ect on the timing and the amplitude of some peaks of the ERPs curves. Functional ANOVA is an adaptation of linear model or analysis of variance to analyse functional observations. The changes in the functional of interest e ects are generally described using smoothing splines. Eilers and Marx proposed to work with P-splines, a combination of B-splines and di erence penalties on coe cients. We de ne a Psplines model for ERPs curves combined with random e ects. In particular, we show that it is a useful alternative to classical strategies requiring the visual and usually imprecise localization of speci c ERP peaks from curves with a low signal-to-noise ratio. [less ▲] Detailed reference viewed: 26 (2 ULg)Discrimination of shifts in a soil microbial community assosciated with TNT-contamination using a functional ANOVA of 16S rRNA hybridized to oligonucleotide microarrays ; ; et al in Environmental Science & Technology (2006), 40 A functional ANOVA analysis of the thermal dissociation of RNA hybridized to DNA microarrays was used to improve discrimination between two soil microbial communities. Following hybridization of in vitro ... [more ▼] A functional ANOVA analysis of the thermal dissociation of RNA hybridized to DNA microarrays was used to improve discrimination between two soil microbial communities. Following hybridization of in vitro transcribed 16S rRNA derived from uncontaminated and 2,4,6- trinitrotoluene contaminated soils to an oligonucleotide microarray containing group- and species-specific perfect match (PM) probes and mismatch (MM) variants, thermal dissociation was used to analyze the nucleic acid bound to each PM-MM probe set. Functional ANOVA of the dissociation curves generally discriminated PM-MM probe sets when Td values (temperature at 50% probe-target dissociation) could not. Maximum discrimination for many PM and MM probes often occurred at temperatures greater than theTd. Comparison of signal intensities measured prior to dissociation analysis from hybridizations of the two soil samples revealed significant differences in domain-, group-, and species-specific probes. Functional ANOVA showed significantly different dissociation curves for 11 PM probes when hybridizations from the two soil samples were compared, even though initial signal intensities for 3 of the 11 did not vary. [less ▲] Detailed reference viewed: 13 (1 ULg)Constraints on concordance measures in bivariate discrete data ; Lambert, Philippe in Journal of Multivariate Analysis (2005), 93 This paper aims to investigate the constraints on dependence measures based on the concept of concordance when discrete random variables are involved. The main technical argument consists in a continuous ... [more ▼] This paper aims to investigate the constraints on dependence measures based on the concept of concordance when discrete random variables are involved. The main technical argument consists in a continuous extension of integer-valued random variables by convolution with unit support kernels [less ▲] Detailed reference viewed: 21 (4 ULg)Local dependence estimation using semi-parametric Archimedean copulas ; Lambert, Philippe in Canadian Journal of Statistics = Revue Canadienne de Statistique (2005), 33 The authors de¯ne a new semiparametric Archimedean copula family having a °exible depen- dence structure. The family's generator is a local interpolation of existing generators. It has locally-de¯ned ... [more ▼] The authors de¯ne a new semiparametric Archimedean copula family having a °exible depen- dence structure. The family's generator is a local interpolation of existing generators. It has locally-de¯ned dependence parameters. The authors present a penalized constrained least-squares method to estimate and smooth these parameters. They illustrate the °exibility of their dependence model in a bivariate survival example. [less ▲] Detailed reference viewed: 18 (0 ULg)ffects of oligofructose on glucose and lipid metabolism in patients with nonalcoholic steatohepatitis : results of a pilot study ; ; Lambert, Philippe et al in European Journal of Clinical Nutrition (2005), 59 Objective: In experimental animals, recent results suggest that the addition of inulin-type fructans such as oligofructose (OFS) in the diet decreases triacylglycerol accumulation in the liver tissue ... [more ▼] Objective: In experimental animals, recent results suggest that the addition of inulin-type fructans such as oligofructose (OFS) in the diet decreases triacylglycerol accumulation in the liver tissue. Therefore, we have investigated the effect of daily ingestion of OFS in seven patients with nonalcoholic steatohepatitis (NASH), confirmed by liver biopsies. Design: They received 16 g/day OFS or maltodextrine (placebo) for 8 weeks in a randomized double-blind crossover design. Energy intake, body composition, liver steatosis and blood parameters were analysed after 4 and 8 weeks of dietary supplementation. Results: Compared to placebo, OFS decreased significantly serum aminotransferases, aspartate aminotransferase after 8 weeks, and insulin level after 4 weeks, but this could not be related to significant effect on plasma lipids. Conclusion: This pilot study supports the putative interest of OFS in the management of liver diseases associated with abnormal lipid accumulation in humans. [less ▲] Detailed reference viewed: 22 (1 ULg)Bayesian proportional hazards model with time varying regression coefficients: a penalized Poisson regression approach Lambert, Philippe ; in Statistics in Medicine (2005), 24 One can fruitfully approach survival problems without covariates in an actuarial way. In narrow time bins, the number of people at risk is counted together with the number of events. The relationship ... [more ▼] One can fruitfully approach survival problems without covariates in an actuarial way. In narrow time bins, the number of people at risk is counted together with the number of events. The relationship between time and probability of an event can then be estimated with a parametric or semi-parametric model. The number of events observed in each bin is described using a Poisson distribution with the log mean speci ed using a exible penalized B-splines model with a large number of equidistant knots. Regression on pertinent covariates can easily be performed using the same log-linear model, leading to the classical proportional hazard model. We propose to extend that model by allowing the regression coe cients to vary in a smooth way with time. Penalized B-splines models will be proposed for each of these coe cients. We show how the regression parameters and the penalty weights can be estimated e ciently using Bayesian inference tools based on the Metropolis-adjusted Langevin algorithm. [less ▲] Detailed reference viewed: 36 (8 ULg)Parametric accelerated failure time models with random effects and an application to kidney transplant survival Lambert, Philippe ; ; et al in Statistics in Medicine (2004), 23 Accelerated failure time models with a shared random component are described, and are used to evaluate the effect of explanatory factors and different transplant centres on survival times following kidney ... [more ▼] Accelerated failure time models with a shared random component are described, and are used to evaluate the effect of explanatory factors and different transplant centres on survival times following kidney transplantation. Different combinations of the distribution of the random effects and baseline hazard function are considered and the fit of such models to the transplant data is critically assessed. A mixture model that combines short- and long-term components of a hazard function is then developed, which provides a more flexible model for the hazard function. The model can incorporate different explanatory variables and random effects in each component. The model is straightforward to fit using standard statistical software, and is shown to be a good fit to the transplant data. [less ▲] Detailed reference viewed: 46 (3 ULg)Analysis of bivariate tail dependence using extreme values copulas: An application to the SOA medical large claims database ; ; Lambert, Philippe in Belgian Actuarial Bulletin (2003) Detailed reference viewed: 41 (4 ULg)Generalized Pareto fit to the Society of Actuaries large claims database ; ; Lambert, Philippe in North American Actuarial Journal (2003), 7 Detailed reference viewed: 55 (1 ULg)Statistical models for the analysis of controlled trials on acute migraine ; Lambert, Philippe in Pharmaceutical Statistics (2003), 2 pecific efficacy criteria were defined by the International Headache Society for controlled clinical trials on acute migraine. They are derived from the pain profile and the timing of rescue medication ... [more ▼] pecific efficacy criteria were defined by the International Headache Society for controlled clinical trials on acute migraine. They are derived from the pain profile and the timing of rescue medication intake. We present a methodology to improve the analysis of such trials. Instead of analysing each endpoint separately, we model the joint distribution and derive success rates in any criteria as predictions. We use cumulative regression models for each response at a time and a multivariate normal copula to model the dependence between responses. Parameters are estimated using maximum likelihood. Benefits of the method include a reduction in the number of tests performed and an increase in their power. The method is well suited to dose-response trials from which predictions can be used to select doses and optimize the design of subsequent trials. More generally, our method permits a very flexible modelling of longitudinal series of ordinal data. [less ▲] Detailed reference viewed: 19 (3 ULg)Improved rank-based dependence measures for categorical data ; Lambert, Philippe in Statistics & Probability Letters (2003), 63 We extend rank-based dependence measures like Spearman's rho to categorical data so that the same ±1 limits are always reached under complete dependence. A goodness-of-fit procedure is derived for ... [more ▼] We extend rank-based dependence measures like Spearman's rho to categorical data so that the same ±1 limits are always reached under complete dependence. A goodness-of-fit procedure is derived for dependence models using copulas. [less ▲] Detailed reference viewed: 19 (1 ULg)A mixture model to assess the effect of a hormonal stimulation on the development of follicles in prepubertal heifers Lambert, Philippe in Journal of the Royal Statistical Society : Series C (Applied Statistics) (2002), 4 The goal of this study is to assess the effect of follicle stimulating hormone (FSH) on the growth of follicles in prepubertal heifers. Adams and co-workers have shown that follicles develop in waves as ... [more ▼] The goal of this study is to assess the effect of follicle stimulating hormone (FSH) on the growth of follicles in prepubertal heifers. Adams and co-workers have shown that follicles develop in waves as in post-pubertal heifers: after a few days, a dominant follicle arises on one of the two ovaries whereas the other subordinate follicles regress.The dominant follicle remains detectable when the next wave occurs. The number and the diameters of these follicles have been measured for 14 days.The same experiment was repeated after a 3-day FSH stimulation. Several large follicles now arise simultaneously on both ovaries together with small follicles which cannot be detected any more when the next wave occurs. Unfortunately, follicles were not individually tracked. Therefore a three-component mixture model was used to describe the potential simultaneous observation of dominant, decaying dominant and subordinate follicles before the FSH stimulation; the same type of model was considered for large, decaying large and small follicles after the hormonal stimulation. Moreover, for technical reasons, only follicles larger than 2 mm are observable. Measurements were thus left truncated.The models fitted are used to quantify the effects of the stimulation and to determine the ideal time point at which follicles should be punctured under ultrasonography to harvest ovocytes that will be fertilized and cultured in vitro. [less ▲] Detailed reference viewed: 8 (1 ULg)A copula based model for multivariate non normal longitudinal data: analysis of a dose titration safety study on a new antidepressant Lambert, Philippe ; in Statistics in Medicine (2002), 21 A new model for multivariate non-normal longitudinal data is proposed. In a first step, each longitudinal series of data corresponding to a given response is modelled separately using a copula to relate ... [more ▼] A new model for multivariate non-normal longitudinal data is proposed. In a first step, each longitudinal series of data corresponding to a given response is modelled separately using a copula to relate the marginal distributions of the response at each time of observation. In a second step, at each observation time, the conditional (on the past) distributions of each response are related using another copula describing the relationship between the corresponding variables. Note that there is no need to consider the same family of distributions for these response variables. The technique is illustrated in a dose titration safety study on a new antidepressant. The haemodynamic effect on diastolic blood pressure, systolic blood pressure and heart rate is studied. These three responses are measured repeatedly over time on ten healthy volunteers during the dose escalation. The available covariates are sex and the concentration of drug in the plasma at time of measurement. [less ▲] Detailed reference viewed: 39 (3 ULg)On the joint analysis of longitudinal responses and early discontinuation in randomized trials ; Lambert, Philippe in Journal of Biopharmaceutical Statistics (2002), 12 Our focus is on the joint analysis of longitudinal nonnormal responses and early discontinuation in (pre)-clinical trials. Separate models are fitted to the two series (response and discontinuation) to ... [more ▼] Our focus is on the joint analysis of longitudinal nonnormal responses and early discontinuation in (pre)-clinical trials. Separate models are fitted to the two series (response and discontinuation) to account for covariate and time effects. The serial dependence and the dependence between response and drop-out are also modeled. This is done using particular dependence functions, called copulas. Copulas are used to create a joint distribution with given marginal distributions. Applications are given for the analysis of heart rate/morbidity in toxicology and pain severity/intake of rescue medications in a trial on migraine. Using copulas, the level of dependence between two variables remains invariant to changes in the marginal distribution of either variable. This proves interesting in modeling the association in a longitudinal setting when responses change over time. [less ▲] Detailed reference viewed: 12 (1 ULg)Effect of conventional controlled-rate freezing and vitrification on quality of bovine blasotcysts produced in vitro ; ; Lambert, Philippe et al in Biology of Reproduction (2001), 65 This study compares the effects of conventional controlled-rate freezing and vitrification on the morphology and metabolism of in vitro-produced bovine blastocysts. Day 7 expanded blastocysts cultured in ... [more ▼] This study compares the effects of conventional controlled-rate freezing and vitrification on the morphology and metabolism of in vitro-produced bovine blastocysts. Day 7 expanded blastocysts cultured in synthetic oviduct fluid with 5% fetal calf serum were frozen in 1.36 M glycerol, 0.25 M sucrose or vitrified in 25% glycerol, 25% ethylene glycol. Cell alterations and in vitro development were evaluated immediately after thawing or after 72 h. The effect of cryopreservation on inner cell mass and trophectoderm (TE) cell number as well as glucose, pyruvate, and oxygen uptakes, and lactate release by blastocysts were evaluated. Immediately after thawing, blastocysts showed equivalent cell membrane permeabilization after both cryopreservation procedures, while alterations in nuclear staining were more frequent in vitrified embryos. After culture, similar survival and hatching rates were observed. Both procedures decreased cell number immediately after thawing and after 72 h. However, the number of TE cells was lower in frozen embryos than in vitrified ones. In relation to this, frozen blastocysts showed a decrease in glucose, pyruvate, and oxygen uptake, although those parameters were not altered in vitrified embryos. An increased glycolytic activity was also observed in frozen embryos, indicating a stress response to this procedure. [less ▲] Detailed reference viewed: 34 (0 ULg)Smoothed nonparametric maximum likelihood estimation of the risk distribution underlying bonus-malus systems ; Lambert, Philippe in Proceedings of the Casualty Actuarial Society (2001), LXXXVIII Mixed Poisson distributions are widely used for modeling claim counts when the portfolio is thought to be heterogeneous. The risk (or mixing) distribution then represents a measure of this heterogeneity ... [more ▼] Mixed Poisson distributions are widely used for modeling claim counts when the portfolio is thought to be heterogeneous. The risk (or mixing) distribution then represents a measure of this heterogeneity. The aim of this paper is to use a variant of the Patilea and Rolin [15] smoothed version of the Simar [20] Non-Parametric Maximum Likelihood Estimator of the risk distribution in the mixed Poisson model. Empirical results based on two data sets from automobile third-party liability insurance demonstrate the relevance of this approach. The design of merit-rating schemes is discussed in the second part of the paper. [less ▲] Detailed reference viewed: 21 (0 ULg)Osmotic behaviour of in vitro produced bovine blastocysts in cryoprotectant solutions as a potential predictive test of survival ; ; Lambert, Philippe et al in Cryobiology (2000), 41 The osmotic behavior of bovine blastocysts produced in vitro was filmed during exposure to and dilution of cryoprotectant solutions used for vitrification. The relationship between the changes in the ... [more ▼] The osmotic behavior of bovine blastocysts produced in vitro was filmed during exposure to and dilution of cryoprotectant solutions used for vitrification. The relationship between the changes in the diameter of embryos and their subsequent survival was assessed. Embryos collected on Day 6 and Day 7 postinsemination were exposed to 10% glycerol (GLY) for 5 min, 10% GLY + 20% ethylene glycol (EG) for 5 min, and 25% Gly + 25% EG for 30 s, before dilution in 0.85 M galactose and finally in embryo transfer freezing medium (ETF). Embryos that had a higher probability of survival behaved as perfect osmometers, shrinking, reexpanding, or swelling according to an identical pattern, whereas embryos that deviated from this standard usually did not survive. The initial embryo diameter, duration of shrinkage and expansion in 10% glycerol, duration of reexpansion in ETF, and final embryo diameter were clearly predictive of the ability to hatch after culture in vitro. On a given day postinsemination, larger blastocysts were more likely than smaller blastocysts to survive and hatch after exposure to cryoprotectants with or without vitrification. [less ▲] Detailed reference viewed: 26 (0 ULg)Alveolar clearance in horses with chronic obstructive pulmonary disease Votion, Dominique ; Vandenput, Sandrina ; et al in American Journal of Veterinary Research (1999), 60 OBJECTIVE: To assess sensitivity of scintigraphic alveolar clearance rate as an indicator of alveolar epithelium damage in horses. ANIMALS: 5 healthy horses (group A) and 5 with chronic obstructive ... [more ▼] OBJECTIVE: To assess sensitivity of scintigraphic alveolar clearance rate as an indicator of alveolar epithelium damage in horses. ANIMALS: 5 healthy horses (group A) and 5 with chronic obstructive pulmonary disease (COPD; group B). PROCEDURE: Horses underwent clearance rate (k [%/min]) determination. Clearance rate of group-B horses was determined after remission of the disease following 2 months at pasture (remission 1), stabling in a controlled environment (remission 2), and during crisis induced by exposure to moldy hay and straw. Methacholine challenge test was performed at each investigation period to determine nonspecific pulmonary airway hyperresponsiveness. Pulmonary function tests (PFT) also were performed, and cell populations in bronchoalveolar lavage (BAL) fluid were determined on another occasion. RESULTS: Group-B horses had significantly faster mean clearance rate during crisis (k = 4.30+/-0.95%/min), compared with that for remission 1(k = 1.98+/-0.55%/min), which did not differ from the rate in group-A horses (k = 1.95+/-0.33%/min). Despite lack of clinical signs of COPD during remission when stabled in a controlled environment, an intermediate value was found (k = 3.20+/-0.72%/min). CONCLUSIONS: This technique allowed grading of lung damage induced by COPD, whereas use of PFT and determination of BAL fluid cell populations failed to differentiate between remission 1 and remission 2. CLINICAL RELEVANCE: Determination of alveolar clearance rate by use of scintigraphy is a sensitive indicator of lung damage. A modified clearance rate was found despite the lack of clinical and functional changes [less ▲] Detailed reference viewed: 63 (18 ULg) |
||