SVM-Maj: a majorization approach to linear support vector machines with different hinge errors

Support vector machines (SVM) are becoming increasingly popular for the prediction of a binary dependent variable. SVMs perform very well with respect to competing techniques. Often, the solution of an SVM is obtained by switching to the dual. In this paper, we stick to the primal support vector machine problem, study its effective aspects, and propose varieties of convex loss functions such as the standard for SVM with the absolute hinge error as well as the quadratic hinge and the Huber hinge errors. We present an iterative majorization algorithm that minimizes each of the adaptations. In addition, we show that many of the features of an SVM are also obtained by an optimal scaling approach to regression. We illustrate this with an example from the literature and do a comparison of different methods on several empirical data sets.     

Distance measurements in the borderline region of applicability of CW EPR and DEER: a model study on a homologous series of spin-labelled peptides

Inter-spin distances between 1 nm and 4.5 nm are measured by continuous wave (CW) and pulsed electron paramagnetic resonance (EPR) methods for a series of nitroxide-spin-labelled peptides. The upper distance limit for measuring dipolar coupling by the broadening of the CW spectrum and the lower distance limit for the present optimally-adjusted double electron electron resonance (DEER) set-up are determined and found to be both around 1.6-1.9 nm. The methods for determining distances and corresponding distributions from CW spectral line broadening are reviewed and further developed. Also, the work shows that a correction factor is required for the analysis of inter-spin distances below approximately 2 nm for DEER measurements and this is calculated using the density matrix formalism.