On comparing,classifying and clustering several dependent regression models

Classifying several regression models fitted on a dataset is one of the most problems in data analysis. In other words, scientists are interested in comparing several regression models that can be used for a dataset. In this paper, an approach will be used to compare and classify several dependent regression models. Then the performance of the proposed method is investigated using simulation study and real example.     

The influence function of the optimal bandwidth for kernel density estimation

Theories about the bandwidth of kernel density estimation have been well established by many statisticians. However, the influence function of the bandwidth has not been well investigated. The influence function of the optimal bandwidth that minimizes the mean integrated square error is derived and the asymptotic property of the bandwidth selectors based on the influence function is provided.     

Measuring Loneliness Among Middle-Aged and Older Adults: The UCLA and de Jong Gierveld Loneliness Scales

This study examined the measurement and invariance properties of the R-UCLA and de Jong Gierveld loneliness scales for research involving middle-aged and older adults. Confirmatory factor analyses (CFAs) of data from interviews with adults aged 45–84 revealed limited support for the unidimensionality of either scale while subsequent analyses indicated the multidimensionality of both. However, method effects associated with positive and negative item wording were also evident. Multiple group CFAs provided limited support for assumptions of measurement invariance across age groups and from baseline to follow-up with regard to the R-UCLA scale. In contrast, strong measurement invariance across age groups and successive measurements was established for the bi-dimensional dJG scale. Overall, the findings supported the relative utility of the dJG scale for research involving middle-aged and older adults but suggested a need for attention to the implications of method effects associated with item wording and lack of measurement invariance with respect to item residuals.     

Affective machines or the inner self? Drawing the boundaries of the female body in the socialist romantic imagination

This article explores two conflicting aesthetics of the female body in post-Stalinist Soviet science fiction. One represented women of the communist future as explorers of the space frontier in assemblages with machines, testing the cultural border between the female and the technological. Another appealed to the mysterious female nature as the Other of human culture, pushing forward the understanding of socialist progress as a masculine project. This article argues that both aesthetics grew within the cultural phenomenon of socialist Romanticism, which emerged in the mid-1950s as a reaction to Stalin-era quasi-Enlightenment rationality and its dominant style of socialist realism.     

Generalized estimating equations for mixtures with varying concentrations

A finite mixture model is considered in which the mixing probabilities vary from observation to observation. A parametric model is assumed for one mixture component distribution, while the others are nonparametric nuisance parameters. Generalized estimating equations (GEE) are proposed for the semi‐parametric estimation. Asymptotic normality of the GEE estimates is demonstrated and the lower bound for their dispersion (asymptotic covariance) matrix is derived. An adaptive technique is developed to derive estimates with nearly optimal small dispersion. An application to the sociological analysis of voting results is discussed. The Canadian Journal of Statistics 41: 217–236; 2013 © 2013 Statistical Society of Canada     

An intuitive risk factors search algorithm: usage of the Bayesian network technique in personalized medicine

The article focuses on the application of the Bayesian networks (BN) technique to problems of personalized medicine. The simple (intuitive) algorithm of BN optimization with respect to the number of nodes using naive network topology is developed. This algorithm allows to increase the BN prediction quality and to identify the most important variables of the network. The parallel program implementing the algorithm has demonstrated good scalability with an increase in the computational cores number, and it can be applied to the large patients database containing thousands of variables. This program is applied for the prediction for the unfavorable outcome of coronary artery disease (CAD) for patients who survived the acute coronary syndrome (ACS). As a result, the quality of the predictions of the investigated networks was significantly improved and the most important risk factors were detected. The significance of the tumor necrosis factor-alpha gene polymorphism for the prediction of the unfavorable outcome of CAD for patients survived after ACS was revealed for the first time.     

Nonparametric estimation for compound Poisson process via variational analysis on measures

The paper develops new methods of nonparametric estimation of a compound Poisson process. Our key estimator for the compounding (jump) measure is based on series decomposition of functionals of a measure and relies on the steepest descent technique. Our simulation studies for various examples of such measures demonstrate flexibility of our methods. They are particularly suited for discrete jump distributions, not necessarily concentrated on a grid nor on the positive or negative semi-axis. Our estimators also applicable for continuous jump distributions with an additional smoothing step.     

The conditional maximum of Poisson random variables

The conditional maxima of independent Poisson random variables are studied. A triangular array of row-wise independent Poisson random variables is considered. If condition is given for the row-wise sums, then the limiting distribution of the row-wise maxima is concentrated onto two points. The result is in accordance with the classical result of Anderson. The case of general power series distributions is also covered. The model studied in Theorems 2.1 and 2.2 is an analogue of the generalized allocation scheme. It can be considered as a non homogeneous generalized scheme of allocations of at most n balls into N boxes. Then the maximal value of the contents of the boxes is studied.     

Testing the equality of two independent regression models

In some situations, for example in agriculture, biology, hydrology, and psychology, researchers wish to determine whether the relationship between response variable and predictor variables differs in two populations. In other words, we are interested in comparing two regression models for two independent datasets. In this work, we will use the parametric and nonparametric methods to establish hypothesis testing for the equality of two independent regression models. Then the simulation study is provided to investigate the performance of the proposed method.