Different kinds of sufficiency in the general Gauss-Markov model
Directional quantile regression in Octave (and MATLAB)
Although many words have been written about two recent directional (regression) quantile concepts, their applications, and the algorithms for computing associated (regression) quantile regions, their software implementation is still not widely available, which, of course, severely hinders the dissemination of both methods. Wanting to partly fill in the gap here, we provide all the codes needed for computing and plotting the multivariate (regression) quantile regions in Octave and MATLAB, describe...
Directional quantile regression in R
Recently, the eminently popular standard quantile regression has been generalized to the multiple-output regression setup by means of directional regression quantiles in two rather interrelated ways. Unfortunately, they lead to complicated optimization problems involving parametric programming, and this may be the main obstacle standing in the way of their wide dissemination. The presented R package modQR is intended to address this issue. It originates as a quite faithful translation of the authors'...
Distribution of estimators and optimal experimental design in nonlinear regression
Distribution of the weighted L.S. estimates in nonlinear models with symmetrical errors
Dynamics of an artificial slope
The slope shape is replaced by a 3D regression function which corresponds with high precision to the position of several hundred points which were determined on the surface of the slope body. The position of several points was repeatedly measured for several years. The time changes in the position of these points were used to create regression functions that describe vertical movements, slope settlement and horizontal movements, slope movement. The model results are presented in the form of mathematical...
Effet des erreurs de mesure sur l'évaluation de l'efficacité d'un traitement
L’effet d’un traitement sur une compétence peut être exprimé par le coefficient de régression partiel avec contrôle de la compétence initiale . Quand et sont mesurées avec erreurs par et , cet effet se manifeste par le coefficient dans la régression de sur et . Le biais entre et est explicité, discuté et montré systématique si et sont corrélés. L’importance de bien spécifier le modèle, dont une condition nécessaire et suffisante d’identification est donnée, est mise en...
Efficient robust estimation of time-series regression models
The paper studies a new class of robust regression estimators based on the two-step least weighted squares (2S-LWS) estimator which employs data-adaptive weights determined from the empirical distribution or quantile functions of regression residuals obtained from an initial robust fit. Just like many existing two-step robust methods, the proposed 2S-LWS estimator preserves robust properties of the initial robust estimate. However, contrary to the existing methods, the first-order asymptotic behavior...
Ein Beitrag zur Analyse von Zeitreihen: Die Quotientenmethode, eine Variante der "Variate Difference" Methode
Eine Bemerkung über das Verhalten der Verbesserungen und ein neuer Vorschlag zur Gewichtung der Meßwerte bei Ausgleichung bedingter direkter Beobachtungen.
El modelo lineal sin término independiente y el coeficiente de determinación. Un estudio Monte Carlo.
En el presente trabajo se analiza y compara mediante un experimento Monte Carlo el comportamiento de cinco expresiones para el Coeficiente de Determinación cuando el modelo lineal se especifica sin término independiente. Se ensayan distintos valores del parámetro poblacional P2, que mide la proporción de varianza explicada por el modelo, introduciendo también la multicolinealidad como factor de variación en el diseño. Se confirma el coeficiente propuesto por Heijmans y Neudecker (1987) y el de Barten...
El sesgo condicionado en el análisis de influencia: una revisión.
El sesgo condicionado se ha propuesto como diagnóstico de influencia en distintos modelos y técnicas estadísticas. Tratando de recoger una visión global de la utilidad del concepto, en este trabajo se hace una revisión general del mismo relacionándolo con la curva de sensibilidad y la curva de influencia muestral. Además, se señalan posibles líneas de trabajo que permitirán abordar el análisis de la influencia a través de este enfoque en una gran variedad de técnicas estadísticas.
Eliminating transformations for nuisance parameters in linear regression models with type I constraints
The linear regression model in which the vector of the first order parameter is divided into two parts: to the vector of the useful parameters and to the vector of the nuisance parameters is considered. The type I constraints are given on the useful parameters. We examine eliminating transformations which eliminate the nuisance parameters without loss of information on the useful parameters.
Eliminating transformations for nuisance parameters in linear model
Elimination of nuisance parameters in a regression model
Empirical regression quantile processes
We address the problem of estimating quantile-based statistical functionals, when the measured or controlled entities depend on exogenous variables which are not under our control. As a suitable tool we propose the empirical process of the average regression quantiles. It partially masks the effect of covariates and has other properties convenient for applications, e.g. for coherent risk measures of various types in the situations with covariates.
Equivalent algorithms for estimation in linear model with conditions
Ergänzung zu Â?Ein Beitrag zur Analyse von Zeitreihen: Die Quotientenmethode, eine Variante der 'Variate Difference' Methode", Metrika, Vol. 16, 1970, Fasc. 2 - 3
Erratum «Deux méthodes linéaires en statistique multidimensionnelle»
Erratum: Second-order approximation of the entropy in nonlinear least-squares estimation