Consider a system of many components with constant failure rate andgeneral repair rate. When all components are reliable and easily reparable,the reliability of the system can be evaluated from the probability q offailure before restoration. In [14], authors give an asymptoticapproximation by monotone sequences. In the same framework, we propose,here, a bounding for q and apply it in the ageing property case.

We consider, in the continuous time version, γ independent random walks on Z+ in random environment in Sinai’s regime. Let Tγ be the first meeting time of one pair of the γ random walks starting at different positions. We first show that the tail of the quenched distribution of Tγ, after a suitable rescaling, converges in probability, to some functional of the Brownian motion. Then we compute the law of this functional. Eventually, we obtain results about the moments of this meeting time. Being Eω the quenched expectation, we show that, for almost all environments ω, Eω[Tγc] is finite for c < γ(γ − 1) / 2 and infinite for c > γ(γ − 1) / 2.

We propose some construction of enhanced Gaussian processes usingKarhunen-Loeve expansion. We obtain a characterization and somecriterion of existence and uniqueness. Using rough-path theory, wederive some Wong-Zakai Theorem.

We present several functional inequalitiesfor finite difference gradients, such asa Cheeger inequality, Poincaré and (modified) logarithmic Sobolev inequalities,associated deviation estimates,and an exponential integrability property.In the particular case of the geometric distribution on ${\mathbb{N}}$ we use an integration by parts formula to computethe optimal isoperimetric and Poincaré constants,and to obtain an improvement of ourgeneral logarithmic Sobolev inequality.By a limiting procedure we recover the correspondinginequalities for the exponential distribution.These results have applications to interacting spin systems undera geometric reference measure.


Using probabilistic tools, this work states a pointwise convergence offunction solutions of the 2-dimensional Boltzmann equation to the functionsolution of the Landau equation for Maxwellian molecules when the collisions become grazing. To this aim, we use the results ofFournier (2000) on the Malliavin calculus for the Boltzmannequation. Moreover, using the particle system introduced by Guérin andMéléard (2003), some simulations of the solution of the Landau equation will be given. This result isoriginal and has not been obtained for the moment by analytical methods.

In a general model (AIMD) of transmission control protocol (TCP)used in internet traffic congestion management, the time dependentdata flow vector x(t) > 0 undergoes a biased random walk ontwo distinct scales. The amount of data of each component x i (t)goes up to x i (t)+a with probability 1-ζ i (x) ona unit scale or down to γ x i (t), 0 < γ < 1 withprobability ζ i (x) on a logarithmic scale, where ζ i depends on the joint state of the system x. Weinvestigate the long time behavior, mean field limit, and the oneparticle case. According to c = lim inf|x|→∞ |x|ζ i (x), the process drifts to ∞ in thesubcritical c < c + (n, γ) case and has an invariantprobability measure in the supercritical case c > c + (n, γ).Additionally, a scaling limit is proved when ζ i (x)and a are of order N –1 and t → N t, in the form of acontinuum model with jump rate α(x).

We consider representations of a joint distribution law of a family of categorical randomvariables (i.e., a multivariate categorical variable) as a mixture ofindependent distribution laws (i.e. distribution laws according to whichrandom variables are mutually independent). For infinite families of random variables, wedescribe a class of mixtures with identifiable mixing measure. This class is interestingfrom a practical point of view as well, as its structure clarifies principles of selectinga “good” finite family of random variables to be used in applied research. For finitefamilies of random variables, the mixing measure is never identifiable; however, it alwayspossesses a number of identifiable invariants, which provide substantial informationregarding the distribution under consideration.

We consider a finite mixture of Gaussian regression models for high-dimensionalheterogeneous data where the number of covariates may be much larger than the sample size.We propose to estimate the unknown conditional mixture density by anℓ1-penalized maximum likelihood estimator. We shall providean ℓ1-oracle inequality satisfied by this Lasso estimator withthe Kullback–Leibler loss. In particular, we give a condition on the regularizationparameter of the Lasso to obtain such an oracle inequality. Our aim is twofold: to extendthe ℓ1-oracle inequality established by Massart and Meynet[12] in the homogeneous Gaussian linearregression case, and to present a complementary result to Städler et al.[18], by studying the Lasso for itsℓ1-regularization properties rather than considering it as avariable selection procedure. Our oracle inequality shall be deduced from a finite mixtureGaussian regression model selection theorem for ℓ1-penalizedmaximum likelihood conditional density estimation, which is inspired from Vapnik’s methodof structural risk minimization [23] and from thetheory on model selection for maximum likelihood estimators developed by Massart in [11].

Initially motivated by a practical issue in target detection vialaser vibrometry, we are interested in the problem of periodicsignal detection in a Gaussian fixed design regression framework.Assuming that the signal belongs to some periodic Sobolev ball andthat the variance of the noise is known, we first consider theproblem from a minimax point of view: we evaluate the so-calledminimax separation rate which corresponds to the minimall 2-distance between the signal and zero so that the detection ispossible with prescribed probabilities of error. Then, we propose atesting procedure which is available when the variance of the noiseis unknown and which does not use any prior information about thesmoothness degree or the period of the signal. We prove that it isadaptive in the sense that it achieves, up to a possible logarithmicfactor, the minimax separation rate over various periodic Sobolevballs simultaneously. The originality of our approach as compared torelated works on the topic of signal detection is that our testingprocedure is sensitive to the periodicity assumption on the signal.A simulation study is performed in order to evaluate the effect ofthis prior assumption on the power of the test. We do observe thegains that we could expect from the theory. At last, we turn to theapplication to target detection by laser vibrometry that we had inview.

This paper derives an explicit approximation for the tail probability of a sum of samplevalues taken without replacement from an unrestricted finite population. The approximationis shown to hold under no conditions in a wide range with relative error given in terms ofthe standardized absolute third moment of the population, β3N. This approximation is used to obtaina result comparable to the well-known Cramér large deviation result in the independentcase, but with no restrictions on the sampled population and an error term depending onlyon β3N. Application to permutation tests isinvestigated giving a new limit result for the tail conditional probability of thestatistic given order statistics under mild conditions. Some numerical results are givento illustrate the accuracy of the approximation by comparing our results to saddlepointapproximations requiring strong conditions.

In this paper we solve the basic fractional analogue of the classical linear-quadratic Gaussian regulator problem in continuous-time with partial observation. For a controlled linear system where both the state and observation processes are driven by fractional Brownian motions, we describe explicitly the optimal control policy which minimizes a quadratic performance criterion. Actually, we show that a separation principle holds, i.e., the optimal control separates into two stages based on optimal filtering of the unobservable state and optimal control of the filtered state. Both finite and infinite time horizon problems are investigated.

We establish some error estimates for the approximation of anoptimal stopping problem along the paths of the Black–Scholesmodel. This approximation is based on a tree method. Moreover, wegive a global approximation result for the related obstacleproblem.

We give a stochastic expansion for estimates $\widehat{\theta}$that minimise the arithmetic mean of (typically independent) random functions of a known parameter θ.Examples include least squares estimates, maximum likelihood estimates and more generally M-estimates.This is used to obtain leading cumulant coefficients of $\widehat{\theta}$needed for the Edgeworth expansions for the distribution and density n1/2 (of \widehat{\theta}$ − θ0) to magnitude n−3/2 (or to n−2 for the symmetric case),where θ0 is the true parameter value and n is typically the sample size.Applications are given to least squares estimates for both real and complex models.An alternative approach is given when the linear parameters of the model are nuisance parameters.The methods are illustrated with the problem of estimating the frequencieswhen the signal consists of the sum of sinusoids of unknown amplitudes.

In this paper, under the linear regression model with heteroscedastic and/or correlated errors when the stochastic linear restrictions on the parameter vector are assumed to be held, a generalization of the ordinary mixed estimator (GOME), ordinary ridge regression estimator (GORR) and Generalized least squares estimator (GLSE) is proposed. The performance of this new estimator against GOME, GORR, GLS and the stochastic restricted Liu estimator (SRLE) [Yang and Xu, Statist. Papers 50 (2007) 639–647] are examined in terms of matrix mean square error criterion. A numerical example is considered to illustrate the theoretical results.

We define multifractional Brownian fields indexed by a metric space, such as a manifold with its geodesic distance, when the distance is of negative type. This construction applies when the Brownian field indexed by the metric space exists, in particular for spheres, hyperbolic spaces and real trees.

A sample of i.i.d. continuous time Markov chains beingdefined, the sum over each component of a real function of thestate is considered. For this functional, a central limit theorem for the first hitting time of a prescribed level is proved. The result extends the classical central limit theorem for order statistics. Various reliability models are presented as examples of applications.

Following the recent investigations of Baik and Suidan in [Int. Math. Res. Not. (2005) 325–337] and Bodineau and Martin in [Electron. Commun. Probab. 10 (2005) 105–112 (electronic)], we prove large deviation properties for a last-passage percolation model in ℤ+2 whose paths are close to the axis. The results are mainly obtained when the random weights are Gaussian or have a finite moment-generating function and rely, as in [J. Baik and T.M. Suidan, Int. Math. Res. Not. (2005) 325–337] and [T. Bodineau and J. Martin, Electron. Commun. Probab. 10 (2005) 105–112 (electronic)], on an embedding in Brownian paths and the KMT approximation. The study of the subexponential case completes the exposition.

We study the large deviation principle for stochastic processes of the form $\{\sum_{k=1}^{\infty}x_{k}(t)\xi_{k}:t\in T\}$ , where $\{\xi_{k}\}_{k=1}^{\infty}$ is a sequence of i.i.d.r.v.'s with mean zero and $x_{k}(t)\in \mathbb{R}$ . We present necessary and sufficient conditions for the large deviation principle for these stochastic processes in several situations. Our approach is based in showing the large deviation principle of the finite dimensional distributions and an exponential asymptotic equicontinuity condition. In order to get the exponential asymptotic equicontinuity condition, we derive new concentration inequalities, which are of independent interest.

Consider an infinite dimensionaldiffusion process process on TZd, whereT is the circle, defined by the action of its generatorL on C2(TZd) local functions as$Lf(\eta)=\sum_{i\in{\bf Z}^d}\left(\frac{1}{2}a_i\frac{\partial^2 f}{\partial \eta_i^2}+b_i\frac{\partial f}{\partial \eta_i}\right)$. Assume that the coefficients, ai and bi are smooth,bounded, finite range with uniformly bounded second orderpartial derivatives, that ai is only a function of $\eta_i$and that $\inf_{i,\eta}a_i(\eta)>0$.Suppose ν is an invariant product measure. Then,if ν is the Lebesgue measure or if d=1,2, it is theunique invariant measure. Furthermore, if ν istranslation invariant, then it is the unique invariant, translationinvariant measure. Now, consider an infinite particlespin system, with state space {0,1}Zd, definedby the action of its generator on local functions f by$Lf(\eta)=\sum_{x\in{\bf Z}^d}c(x,\eta)(f(\eta^x)-f(\eta))$,where $\eta^x$ is the configuration obtained from ηaltering only the coordinate at site x. Assume that $c(x,\eta)$are of finite range, bounded and that $\inf_{x,\eta}c(x,\eta)>0$.Then, if ν is an invariant product measure for this process,ν is unique when d=1,2. Furthermore, if ν is translationinvariant, it is the unique invariant, translation invariantmeasure. The proofs of these results show how elementarymethods can give interesting information for general processes.

We study asymptotic behavior of Markov chain Monte Carlo (MCMC) procedures. Sometimes theperformances of MCMC procedures are poor and there are great importance for the study ofsuch behavior. In this paper we call degeneracy for a particular type of poorperformances. We show some equivalent conditions for degeneracy. As an application, weconsider the cumulative probit model. It is well known that the natural data augmentation(DA) procedure does not work well for this model and the so-called parameter-expanded dataaugmentation (PX-DA) procedure is considered to be a remedy for it. In the sense ofdegeneracy, the PX-DA procedure is better than the DA procedure. However, when the numberof categories is large, both procedures are degenerate and so the PX-DA procedure may notprovide good estimate for the posterior distribution.