Secondo fascicolo del 2004
Kurtosis Diagram for the Log-Dagum Distribution
by Filippo Domma
The aim of this paper is to study, through the kurtosis diagram proposed by Zenga (1996), the kurtosis of the log-Dagum distribution. In particular, we analyse how changes the kurtosis on the left and right side of the distribution when the parameter β changes, having fixed the median and absolute mean value from the median. We prove that the kurtosis depends only on parameter β and we show that a reduction of the kurtosis on the left side of the diagram and an increse of the kurtosis on the right side of the diagram occours when β increases.
Keywords: Log-Dagum distribution, kurtosis diagram, symmetric and asymmetric distribution.
Random variate generation from
Multivariate Exponential Power distribution
by Nadia Solaro
In this paper two procedures for random variate generation for the family of Multivariate Exponential Power (MEP) distributions are developed and then compared. By conveniently expressing MEP distributions in terms of mixture of random variables, a simple and fast procedure based on transformation approach is proposed. Another procedure, based on a general acceptance-rejection (AR) technique for multivariate orthounimodal densities, is considered. As expected, the transformation-based procedure is considerably less time-consuming than the other one. Furthermore, this study allows us to evaluate to which extent generalizations of AR methods could be computationally less efficient than a transformation one for the type of multivariate distributions considered.
Keywords: acceptance-rejection method, multivariate elliptical and symmetric distributions, multivariate orthounimodal distributions, transformation method.
Ricerca di soglie critiche per un test su matrici di confusione
by Eugenio Brentari, Elena Poli
This paper will discuss misclassification errors induced by human judgement. Facing different classifications coming from different trials, often researchers have to evaluate the goodness of classification proposed. Particularly, we propose an interesting application concerning olfactory measurement. In that field the accuracy of a confusion matrix is crucial, because the classification suggested by a panel of experts is collected by means of a confusion matrix.
The analysis of a confusion matrix we will perform in this study serves both to evaluate skills of judges, of whom the panel is composed, and to evaluate the discriminant capability of the element used to make the experiment.
This work will focus in establishing critical thresholds for the test thorough which we asses not random classification.
Keywords: Confusion Matrix, Misclassification, Critical Thresholds.
Alcune considerazioni sull’efficienza degli stimatori rapporto-cum-prodotto
by Pier Francesco Perri
In order to estimate the unknown mean of a study variable using two auxiliary variables, the efficiency of a modified version of the ratio-cum-product estimators is analysed when the sampling fraction is less then 0.5. Approximate expressions for the bias and the mean square error are obtained and specific conditions under which the modified estimators are more efficient then the original ones are analytically defined. Also a comparison with the multivariate regression estimator is included.
Keywords: auxiliary variables, efficiency, sampling fraction, ratio-cum-product estimators, regression estimator.
L’incidenza dei fattori ambientali sul rischio di malattia: un modello anisotropico
by Eugenia Nissi, Annalina Sarra
In the recent years, there has been growing interest in the development of statistical methods useful in the detection of disease cases related to environmental pollution.
Our starting point is an inhomogeneous Poisson point process model for the spatial variation in the incidence of cases and controls in a designated geographic region, as proposed by Diggle.
In this paper we focus on a particular model with directional effects to analyse the health risk around a point source.
In order to evaluate the adequacy of the proposed model we perform a Monte Carlo simulation study. In particular we analyse the maximum likelihood estimators properties of the model parameters.
Keywords: processi spaziali di punto, regressione binaria, epidemiologia ambientale, modello con effetti direzionali.