On multiple normal probabilities of rectangles
On selecting the best features in a noisy environment
This paper introduces a novel method for selecting a feature subset yielding an optimal trade-off between class separability and feature space dimensionality. We assume the following feature properties: (a) the features are ordered into a sequence, (b) robustness of the features decreases with an increasing order and (c) higher-order features supply more detailed information about the objects. We present a general algorithm how to find under those assumptions the optimal feature subset. Its performance...
On solution sets of information inequalities
We investigate solution sets of a special kind of linear inequality systems. In particular, we derive characterizations of these sets in terms of minimal solution sets. The studied inequalities emerge as information inequalities in the context of Bayesian networks. This allows to deduce structural properties of Bayesian networks, which is important within causal inference.
On the control of the difference between two Brownian motions: a dynamic copula approach
We propose new copulae to model the dependence between two Brownian motions and to control the distribution of their difference. Our approach is based on the copula between the Brownian motion and its reflection. We show that the class of admissible copulae for the Brownian motions are not limited to the class of Gaussian copulae and that it also contains asymmetric copulae. These copulae allow for the survival function of the difference between two Brownian motions to have higher value in the right...
On the control of the difference between two Brownian motions: an application to energy markets modeling
We derive a model based on the structure of dependence between a Brownian motion and its reflection according to a barrier. The structure of dependence presents two states of correlation: one of comonotonicity with a positive correlation and one of countermonotonicity with a negative correlation. This model of dependence between two Brownian motions B1 and B2 allows for the value of [...] to be higher than 1/2 when x is close to 0, which is not the case when the dependence is modeled by a constant...
Overview of Recent Results in Growth-curve-type Multivariate Linear Models
The Extended Growth Curve Model (ECGM) is a multivariate linear model connecting different multivariate regression models in sample subgroups through common variance matrix. It has the form: Here, matrices contain subgroup division indicators, and corresponding regressors. If , we speak about (ordinary) Growth Curve Model. The model has already its age (it dates back to 1964), but it has many important applications. That is why it is still intensively studied. Many articles investigating...
Problèmes engendrés par les données manquantes dans la pratique statistique
Ratio-type Estimatorsfor Multinomial Probabilities in two-phase Sampling
Regularization for high-dimensional covariance matrix
In many applications, high-dimensional problem may occur often for various reasons, for example, when the number of variables under consideration is much bigger than the sample size, i.e., p >> n. For highdimensional data, the underlying structures of certain covariance matrix estimates are usually blurred due to substantial random noises, which is an obstacle to draw statistical inferences. In this paper, we propose a method to identify the underlying covariance structure by regularizing...
Relationship between stochastic inequalities and some classical mathematical inequalities.
Remark on properties of bases for additive logratio transformations of compositional data
The statistical analysis of compositional data, multivariate data when all its components are strictly positive real numbers that carry only relative information and having a simplex as the sample space, is in the state-of-the-art devoted to represent compositions in orthonormal bases with respect to the geometry on the simplex and thus provide an isometric transformation of the data to an usual linear space, where standard statistical methods can be used (e.g. [2], [4], [5], [9]). However, in some...
Remarks on Anděl's paper “On multiple normal probabilities of rectangles”
Robustness regions for measures of risk aggregation
One of risk measures’ key purposes is to consistently rank and distinguish between different risk profiles. From a practical perspective, a risk measure should also be robust, that is, insensitive to small perturbations in input assumptions. It is known in the literature [14, 39], that strong assumptions on the risk measure’s ability to distinguish between risks may lead to a lack of robustness. We address the trade-off between robustness and consistent risk ranking by specifying the regions in...
Segmentation aux moindres carrés : un aspect synthétique
Self-tuning regulators with restricted inputs
Simultaneous rank test procedures
Simultaneous rank test procedures are proposed for testing of randomness concerning some marginals. The considered test procedures are analogous to those introduced by Krishnaiah for classical normal theory (see Krishnaiah (1965) Ann. Inst. Statist. Math. 17, 35-53).
Soft modelling: intermediate between traditional model building and data analysis
Solution of some problems of minimax control for a multivariate linear stochastic system
Strong approximations of bivariate uniform empirical processes
Suitability of linearization of nonlinear problems not only in biology and medicine
Biology and medicine are not the only fields that present problems unsolvable through a linear models approach. One way to overcome this obstacle is to use nonlinear methods, even though these are not as thoroughly explored. Another possibility is to linearize and transform the originally nonlinear task to make it accessible to linear methods. In this aricle I investigate an easy and quick criterion to verify suitability of linearization of nonlinear problems via Taylor series expansion so that...