We establish posterior consistency for non-parametric Bayesian estimation of the dispersion coefficient of a time-inhomogeneous Brownian motion.

Using the concept of Hellinger integrals, necessary and sufficient conditions are established for the contiguity of two sequences of distributions of Poisson point processes with an arbitrary state space. The distribution of logarithm of the likelihood ratio is shown to be infinitely divisible. The canonical measure is expressed in terms of the intensity measures. Necessary and sufficient conditions for the LAN-property are formulated in terms of the corresponding intensity measures.

General conditions for convergence rates of nonparametric orthogonal series estimators of the regression function f(x)=E(Y | X = x) are considered. The estimators are obtained by the least squares method on the basis of a random observation sample (Yi,Xi), i=1,...,n, where $X_i\in A\subset {ℝ}^d$ have marginal distribution with density $\varrho \in L^1\left(A\right)$ and Var( Y | X = x) is bounded on A. Convergence rates of the errors $E_X{(f\left(X\right)-{\widehat{f}}_N\left(X\right))}^2$ and $\parallel f-{\widehat{f}}_N{\parallel }_{\infty }$ for the estimator ${\widehat{f}}_N\left(x\right)={\sum }_{k=1}^N{\widehat{c}}_k{e}_k\left(x\right)$, constructed using an orthonormal system $e_k$, k=1,2,..., in $L^2\left(A\right)$ are obtained.

We propose a feature selection method for density estimation with quadratic loss. This method relies on the study of unidimensional approximation models and on the definition of confidence regions for the density thanks to these models. It is quite general and includes cases of interest like detection of relevant wavelets coefficients or selection of support vectors in SVM. In the general case, we prove that every selected feature actually improves the performance of the estimator. In the case...

This paper is devoted to the study of some asymptotic properties of a $M$-estimator in a framework of detection of abrupt changes in random field’s distribution. This class of problems includes e.g. recovery of sets. It involves various techniques, including $M$-estimation method, concentration inequalities, maximal inequalities for dependent random variables and $\phi$-mixing. Penalization of the criterion function when the size of the true model is unknown is performed. All the results apply under mild,...

This paper is devoted to the study of some asymptotic properties of a M-estimator in a framework of detection of abrupt changes in random field's distribution. This class of problems includes e.g. recovery of sets. It involves various techniques, including M-estimation method, concentration inequalities, maximal inequalities for dependent random variables and ϕ-mixing. Penalization of the criterion function when the size of the true model is unknown is performed. All the results apply under...

The index of regularity of a measure was introduced by Beirlant, Berlinet and Biau [1] to solve practical problems in nearest neighbour density estimation such as removing bias or selecting the number of neighbours. These authors proved the weak consistency of an estimator based on the nearest neighbour density estimator. In this paper, we study an empirical version of the regularity index and give sufficient conditions for its weak and strong convergence without assuming absolute continuity or...

This paper studies quantile linear regression models with response data missing at random. A quantile empirical-likelihood-based method is proposed firstly to study a quantile linear regression model with response data missing at random. It follows that a class of quantile empirical log-likelihood ratios including quantile empirical likelihood ratio with complete-case data, weighted quantile empirical likelihood ratio and imputed quantile empirical likelihood ratio are defined for the regression...

Se definen en este trabajo r-desarrollos de Neumann y se prueba que toda densidad de probabilidad f admite un desarrollo r-convergente a f.Los resultados obtenidos se aplican a la estimación de f sin la suposición de que sea de cuadrado integrable, estudiándose propiedades asintóticas de los estimadores e ilustrándose con un ejemplo de aplicación.

Sea {Xt: t ∈ Z} una serie de tiempo estacionaria, con valores en Rp, verificando la condición de ser α-mixing o L2-estable. A partir de una muestra de tamaño n se define una amplia clase de estimadores no paramétricos de la función de densidad f(x) asociada al proceso, y de la función de autorregresión de orden k:r(y) = E(g(Xt+1)/(Xt-k+1 ... Xt) = y), y ∈ Rksiendo g una función real.Se estudian las siguientes propiedades asintóticas de estos estimadores: consistencia puntual (casi segura y en media...

Sea X una variable aleatoria con función de distribución F(x) y función de densidad f(x) y X1, X2,..., Xn un conjunto de observaciones de la variable que pueden ser dependientes. Se definen dos estimadores no paramétricos generales (uno recursivo y el otro no recursivo) de la función de distribución.Bajo condiciones aceptables se obtiene el sesgo y la varianza y covarianza asintótica de los estimadores definidos. Finalmente se prueban propiedades de consistencia y normalidad asintótica.

Se estudia la estimación de tipo no paramétrico de la función de riesgo o razón de fallo de una variable aleatoria real. A partir de una muestra X1, X2, ..., Xn de datos no censurados y no necesariamente independientes, se considera un estimador cociente entre el estimador núcleo de la función de densidad y un estimador núcleo de la función de supervivencia, sobre el que se estudia el problema de selección del parámetro ventana. Por medio de un estudio de simulación se observa la ventaja de utilizar...

The aim is to study the asymptotic behavior of estimators and tests for the components of identifiable finite mixture models of nonparametric densities with a known number of components. Conditions for identifiability of the mixture components and convergence of identifiable parameters are given. The consistency and weak convergence of the identifiable parameters and test statistics are presented for several models.