Quantile regression with group lasso for classification

Applications of regression models for binary response are very common and models specific to these problems are widely used. Quantile regression for binary response data has recently attracted attention and regularized quantile regression methods have been proposed for high dimensional problems. When the predictors have a natural group structure, such as in the case of categorical predictors converted into dummy variables, then a group lasso penalty is used in regularized methods. In this paper, we present a Bayesian Gibbs sampling procedure to estimate the parameters of a quantile regression model under a group lasso penalty for classification problems with a binary response. Simulated and real data show a good performance of the proposed method in comparison to mean-based approaches and to quantile-based approaches which do not exploit the group structure of the predictors.     

FEM-analysis of a multiferroic nanocomposite – comparison of experiment and numerical simulation

The combination of electric and magnetic materials opens new possibilities in the field of sensor technologies and data storage [1]. These magneto-electric (ME) materials have the property to change a physical ferroic quantity into another, i.e. a magnetic field can change the electric polarization and vice versa. The combination of multiple ferroic characteristics within materials is called multiferroic. Since magneto-electric single-phase materials are rare in nature and typically operate only at very low temperature, they are not favorable in technical applications. However, ME composites, consisting of ferroelectric and ferromagnetic phases, produce a strain-induced magneto-electric product property at room temperature [2]. In these composites, two different effects can be differentiated, the direct and the converse ME effect. The first one describes a polarization which is magnetically caused. In detail, a magnetic field is applied which produces a deformation of the magneto-active phase which is transferred to the electro-active phase and as a consequence this phase exhibits a polarization. Therefore, one can discover a strain-induced polarization. The second effect to observe is a magnetization caused by an electric field. In our contribution, we focus on a (1-3) composite, where cobalt ferrite nanopillars are embedded in a barium titanate matrix, see the experiments described in [3]. In the numerical simulations we compare the changes of the strain-induced inplane polarizations of the ferroelectric matrix with experimental measurements. Furthermore, we analyze the magneto-electric coupling coefficient. (© 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

Vanadium oxide aerogels: Nanostructured materials for enhanced energy storage

The synthesis, chemistry, local structure and electrochemical properties of vanadium oxide xerogels and aerogels have much in common. The one difference in their respective synthesis routes, the means by which solvent is removed, has a significant influence on the resulting morphology. The high surface area, nanodimensional solid phase, short diffusion paths and interconnected mesoporosity of the aerogels exert a profound effect on their electrochemical properties. Our studies with V₂O₅ aerogels show that these materials offer the promise of achieving both high energy density and high power density because of a pseudocapacitive charge storage mechanism which develops.     

The mechanisms of spin delocalization through sigma bonds delocalization to gamma protons [1]

A study of the mechanisms of spin delocalization to γ hydrogens was conducted by investigating the change in spin density distribution upon geometrical deformation. The mechanisms are expressed in valence bond pictures. The spin densities were calculated using the SCF-MO-INDO method for the propyl radical. An empirical expression for γ-hydrogen hfsc's is proposed.