Sample density as the function-estimate of population's distribution
Scatter halfspace depth: Geometric insights
Scatter halfspace depth is a statistical tool that allows one to quantify the fitness of a candidate covariance matrix with respect to the scatter structure of a probability distribution. The depth enables simultaneous robust estimation of location and scatter, and nonparametric inference on these. A handful of remarks on the definition and the properties of the scatter halfspace depth are provided. It is argued that the currently used notion of this depth is well suited especially for symmetric...
Scenario generation with distribution functions and correlations
In this paper, we present a method for generating scenarios for two-stage stochastic programs, using multivariate distributions specified by their marginal distributions and the correlation matrix. The margins are described by their cumulative distribution functions and we allow each margin to be of different type. We demonstrate the method on a model from stochastic service network design and show that it improves the stability of the scenario-generation process, compared to both sampling and a...
Second order asymptotic distribution of the -divergence goodness-of-fit statistics
The distribution of each member of the family of statistics based on the -divergence for testing goodness-of-fit is a chi-squared to (Pardo [pard96]). In this paper a closer approximation to the exact distribution is obtained by extracting the -dependent second order component from the term.
Segmentation of the Poisson and negative binomial rate models: a penalized estimator
We consider the segmentation problem of Poisson and negative binomial (i.e. overdispersed Poisson) rate distributions. In segmentation, an important issue remains the choice of the number of segments. To this end, we propose a penalized -likelihood estimator where the penalty function is constructed in a non-asymptotic context following the works of L. Birgé and P. Massart. The resulting estimator is proved to satisfy an oracle inequality. The performances of our criterion is assessed using simulated...
Selección de la ventana en suavización tipo núcleo de la parte no paramétrica de un modelo parcialmente lineal con errores autorregresivos.
Supongamos que yi = ζiT β + m(ti) + εi, i = 1, ..., n, donde el vector (p x 1) β y la función m(·) son desconocidos, y los errores εi provienen de un proceso autorregresivo de orden uno (AR(1)) estacionario. Discutimos aquí el problema de la selección del parámetro ventana de un estimador tipo núcleo de la función m(·) basado en un estimador Generalizado de Mínimos Cuadrados de β. Obtenemos la expresión asintótica de una ventana óptima y proponemos un método para estimarla, de modo que dé lugar...
Semiparametric deconvolution with unknown noise variance
This paper deals with semiparametric convolution models, where the noise sequence has a gaussian centered distribution, with unknown variance. Non-parametric convolution models are concerned with the case of an entirely known distribution for the noise sequence, and they have been widely studied in the past decade. The main property of those models is the following one: the more regular the distribution of the noise is, the worst the rate of convergence for the estimation of the signal’s density...
Semiparametric deconvolution with unknown noise variance
This paper deals with semiparametric convolution models, where the noise sequence has a Gaussian centered distribution, with unknown variance. Non-parametric convolution models are concerned with the case of an entirely known distribution for the noise sequence, and they have been widely studied in the past decade. The main property of those models is the following one: the more regular the distribution of the noise is, the worst the rate of convergence for the estimation of the signal's density...
Semiparametric estimation in autoregressive processes with Markov regime. (Estimación semiparamétrica en procesos autorregresivos con régimen de Markov.)
Semiparametric estimation of the parameters of multivariate copulas
In the paper we investigate properties of maximum pseudo-likelihood estimators for the copula density and minimum distance estimators for the copula. We derive statements on the consistency and the asymptotic normality of the estimators for the parameters.
Sequential estimation of survival functions with a neutral to the right process prior
In this work, a parametric sequential estimation method of survival functions is proposed in the Bayesian nonparametric context when neutral to the right processes are used. It is proved that the mentioned method is an 1-SLA rule when Dirichlet processes are used; furthermore, asymptotically pointwise optimal procedures are obtained. Finally, an example is given.
Set estimation under convexity type assumptions
Shape factor extremes for prolate spheroids
Microscopic prolate spheroids in a given volume of an opaque material are considered. The extremes of the shape factor of the spheroids are studied. The profiles of the spheroids are observed on a random planar section and based on these observations we want to estimate the distribution of the extremal shape factor of the spheroids. We show that under a tail uniformity condition the Maximum domain of attraction is stable. We discuss the normalising constants (n.c.) for the extremes of the spheroid...
Sharp bounds for expectations of spacings from decreasing density and failure rate families
We apply the method of projecting functions onto convex cones in Hilbert spaces to derive sharp upper bounds for the expectations of spacings from i.i.d. samples coming from restricted families of distributions. Two families are considered: distributions with decreasing density and with decreasing failure rate. We also characterize the distributions for which the bounds are attained.
Sharp bounds on quasiconvex moments of generalized order statistics.
Sharp upper bounds for the best predictor of future mean of some order statistics
We provide sharp upper bounds for the mean of the future order statistics based on observed r order statistics. These bounds are expressed in terms of various scale units. We also determine the probability distributions for which the bounds are attained.
Sign and Wilcoxon tests for quadratic versus cubic regression.
In this paper sign and Wilcoxon tests for testing the null hypothesis of quadratic regression versus the alternative, cubic regression are proposed. It is shown that in the case of a simple design consisting of multiple Y observations at each of the four levels of x, the proposed tests perform reasonably well as compared to their parametric competitors, while in the case of a general design consisting of a large number of levels of x, the loss in Pitman efficiency is considerable. However their...
Significance testing in exact logistic multiple regression.
Simple estimators of the parameters of generalized Tukey’s -family
Simple large sample estimators of scale and location parameters based on blocks of order statistics.
In this paper quite efficient large sample estimation procedures are derived for jointly estimating the parameters of the location-scale family of distributions. These estimators are linear combinations of the means of suitably chosen blocks of order statistics. For specific distributions, such as the extreme-value, normal, and logistic, little is to be gained by using more than three blocks. For these distributions we can obtain joint relative asymptotic efficiencies of 97-98% using the means of...