Feature Selection using Fuzzy Support Vector Machines

The feature selection consists of obtaining a subset of these features to optimally realize the task without the irrelevant ones. Since it can provide faster and cost-effective learning machines and also improve the prediction performance of the predictors, it is a crucial step in machine learning. The feature selection methods using support machines have obtained satisfactory results, but the noises and outliers often reduce the performance. In this paper, we propose a feature selection approach using fuzzy support vector machines and compare it with the previous work, the results of experiments on the UCI data sets show that feature selection using fuzzy SVM obtains better results than using SVM.     

Strength of cylinder joint adhesively bonded by coupling

Summary A method of joining two metal cylindrical shafts with adhesive coupling is proposed. Two cylindrical shafts with the same diameter are connected by bonding through a cylindrical coupling with epoxy resin. The strength of the shaft joint under tensile loading and torsional loading is investigated analytically and experimentally. The stress and strain distributions of the shaft joint is analyzed by the finite element method. The analyzed strain distributions in the joint are compared with experimental values. The joint strength is predicted by applying the strength laws of shafts, coupling, adhesive layer and adhesive interface between shaft and adhesive coupling. The effects of the coupling dimension on the joint strength are examined. It is shown that the adhesive shaft joint can transfer the load by which the cylindrical shafts are plastically deformed.This paper was refined by the author, K. Ikegami, during statying at Technische Universität München under the support of Deutscher Akademischer Austauschdients. The author is grateful to Professor Lippmann of Technische Universität München who is the host professor of the support.     

Hindered motions in a ketonea-urea inclusion compound. An ESR and ENDOR study

We have studied by ESR and ENDOR spectroscopy the free radicals produced in γ-irradiated inclusion compound formed between the ketone 10-nonadecanone and urea. Only one type of long lived radical is formed by the removal of an α-proton from the ketone. The hyperfine (hf) coupling constants of the α- and β-protons of the radicals have been measured by ESR at different temperatures in the range 110–292 K and at different orientations of the crystals. The hf coupling of the γ-protons of the radical and of the urea protons have been studied by ENDOR. The temperature and angular dependences of the coupling constants have been analyzed in terms of the internal and overall motions of the radical inside the hexagonal channels formed by the urea molecules. It has been found that the radical cannot perform complete reorientations around the long molecular axis, but it undergoes restricted rotational diffusion. This process is explained by assuming a coupling between the rotational and translational degrees of freedom of the radical inside the urea channels.     

Stability of bars made of linear elastic materials with voids

Summary Dynamic stability of an elastic bar with voids is considered. Using the Lyapunov approach some new sufficient stability conditions are obtained and explicit expressions for the critical load are derived.     

Medicinal Organometallic Chemistry: Designing Metal Arene Complexes as Anticancer Agents

The field of medicinal inorganic chemistry is rapidly advancing. In particular organometallic complexes have much potential as therapeutic and diagnostic agents. The carbon‐bound and other ligands allow the thermodynamic and kinetic reactivity of the metal ion to be controlled and also provide a scaffold for functionalization. The establishment of structure–activity relationships and elucidation of the speciation of complexes under conditions relevant to drug testing and formulation are crucial for the further development of promising medicinal applications of organometallic complexes. Specific examples involving the design of ruthenium and osmium arene complexes as anticancer agents are discussed.     

Genetically Encoded Bioluminescent Indicator for ERK2 Dimer in Living Cells

In this study, a genetically encoded bioluminescent indicator for ERK2 dimer was developed with the split Renilla luciferase complementation method, in which the formation of ERK2 dimer induces a spontaneous emission of bioluminescence in living cells. In response to extracellular stimuli, such as epidermal growth factor (EGF) or 17β‐estradiol (E2), extracellular signal‐regulated kinase 2 (ERK2) is phosphorylated by its upstream kinase MEK, and also phosphorylates its substrates in various regions of the cell, including the nucleus. Phosphorylated ERK2 is led to form its dimer, thereby transporting itself into the nucleus. We demonstrated with the indicator that stimulation with EGF or E2 induces the formation of ERK2 dimer in living MCF‐7 cells. The dynamics of this dimer formation was examined and discussed.     

Prädikatives und funktionales Denken in der Wahrscheinlichkeitsrechnung

In order to emphasize functional thinking in mathematical education, arguments are brought forward for an early and frequent use of tree diagrams in teaching stochastics. Since set and tree diagrams can be regarded as two sides of the same matter, Bayes’ theorem, which is strongly associated with predicative set diagrams, need not be dealt with explicitly any longer. Set diagrams are replaced by tree diagrams that are labelled in detail and by tree inversion as a functional instrument of dealing with conditioned probabilities. This technique is demonstrated by three typical problems in the context of conditioned probability. Finally, two letters by Pascal to Fermat are analysed in order to illustrate that predicative and functional approaches were already pursued in the early stages of probability calculus in the 17^th century.