B. Lautrup

Measures for measures

Are some ways of measuring scientific quality better than others? Sune Lehmann, Andrew D. Jackson and Benny E. Lautrup analyse the reliability of commonly used methods for comparing citation records.Statistical noiseCitation analysis can loom large in a scientists career. In this issue Sune Lehmann, Andrew Jackson and Benny Lautrup compare commonly used measures of author quality. The mean number of citations per paper emerges as a better indicator than the more complex Hirsch index; a third method, the number of papers published per year, measures industry rather than ability. Careful citation analyses are useful, but Lehmann et al. caution that institutions often place too much faith in decisions reached by algorithm, use poor methodology or rely on inferior data sets.

RECENT DEVELOPMENTS IN THE COMPARISON BETWEEN THEORY AND EXPERIMENTS IN QUANTUM ELECTRODYNAMICS.

Abstract This review is a survey of three main topics in quantum electrodynamics: fundamental bound systems anomalous magnetic moments and high energy experiments. The emphasis lies particularly in recent developments concerninf the electron and muon anomalous magnetic moments.

Protein secondary structure and homology by neural networks The α-helices in rhodopsin

Neural networks provide a basis for semiempirical studies of pattern matching between the primary and secondary structures of proteins. Networks of the perceptron class have been trained to classify the amino‐acid residues into two categories for each of three types of secondary feature: α‐helix or not, β‐sheet or not, and random coil or not. The explicit prediction for the helices in rhodopsin is compared with both electron microscopy results and those of the Chou‐Fasman method. A new measure of homology between proteins is provided by the network approach, which thereby leads to quantification of the differences between the primary structures of proteins.

Phase transition in four-dimensional compact QED

Abstract The energy and the specific heat of the four-dimensional U(1) lattice gauge model is evaluated by Monte Carlo simulations on lattices of size L 4 , where L = 4, 5 and 6, evidence is presented for the occurence of a second-order phase transition. A finite size scaling analysis of our results gives the critical value of the coupling constant e 2 c = 0.995 and a correlation length exponent v ≈ 1 3 .

A novel approach to prediction of the 3-dimensional structures of protein backbones by neural networks

Three‐dimensional structures of protein backbones have been predicted using neural networks. A feed forward neural network was trained on a class of functionally, but not structurally, homologous proteins, using backpropagation learning. The network generated tertiary structure information in the form of binary distance constraints for the Cα atoms in the protein backbone. The binary distance between two Cα atoms was 0 if the distance between them was less than a certain threshold distance, and 1 otherwise. The distance constraints predicted by the trained neural network were utilized to generate a folded conformation of the protein backbone, using a steepest descent minimization approach.

Citation networks in high energy physics.

The citation network constituted by the SPIRES database is investigated empirically. The probability that a given paper in the SPIRES database has k citations is well described by simple power laws, P(k) proportional to k(-alpha), with alpha approximately 1.2 for k less than 50 citations and alpha approximately 2.3 for 50 or more citations. A consideration of citation distribution by subfield shows that the citation patterns of high energy physics form a remarkably homogeneous network. Further, we utilize the knowledge of the citation distributions to demonstrate the extreme improbability that the citation records of selected individuals and institutions have been obtained by a random draw on the resulting distribution.

A quantitative analysis of indicators of scientific performance

Condensing the work of any academic scientist into a one-dimensional indicator of scientific performance is a difficult problem. Here, we employ Bayesian statistics to analyze several different indicators of scientific performance. Specifically, we determine each indicator’s ability to discriminate between scientific authors. Using scaling arguments, we demonstrate that the best of these indicators require approximately 50 papers to draw conclusions regarding long term scientific performance with usefully small statistical uncertainties. Further, the approach described here permits statistical comparison of scientists working in distinct areas of science.Condensing the work of any academic scientist into a one-dimensional measure of scientific quality is a difficult problem. Here, we employ Bayesian statistics to analyze several different measures of quality. Specifically, we determine each measures ability to discriminate between scientific authors. Using scaling arguments, we demonstrate that the best of these measures require approximately 50 papers to draw conclusions regarding long term scientific performance with usefully small statistical uncertainties. Further, the approach described here permits the value-free (i.e., statistical) comparison of scientists working in distinct areas of science.

The stability of solitons in biomembranes and nerves

We examine the stability of a class of solitons, obtained from a generalization of the Boussinesq equation, which have been proposed to be relevant for pulse propagation in biomembranes and nerves. These solitons are found to be stable with respect to small-amplitude fluctuations. They emerge naturally from non-solitonic initial excitations and are robust in the presence of dissipation. Solitary waves pass through each other with only minor dissipation when their amplitude is small. Large-amplitude solitons fall apart into several pulses and small-amplitude noise upon collision when the maximum density of the membrane is limited by the density of the solid phase membrane.

Nonlinear versus Linear Models in Functional Neuroimaging: Learning Curves and Generalization Crossover

We introduce the concept of generalization for models of functional neuroactivation, and show how it is affected by the number, N, of neuroimaging scans available. By plotting generalization as a function of N (i.e. a “learning curve”) we demonstrate that while simple, linear models may generalize better for small Ns, more flexible, low-biased nonlinear models, based on artificial neural networks (ANNs), generalize better for larger Ns. We demonstrate that for sets of scans of two simple motor tasks—one set acquired with [O15]water using PET, and the other using fMRI—practical Ns exist for which “generalization crossover” occurs. This observation supports the application of highly flexible, ANN models to sufficiently large functional activation datasets.

Phase transitions and mean-field methods in lattice gauge theory

Abstract We determine the transition points β c and plaquette energies E p of U(1), SU(2) and SO(3) lattice gauge theories in 4 and 5 dimensions by mean-field methods, and compare with corresponding values determined by the Monte Carlo technique. In every case, the mean-field and Monte Carlo values for β c and E p agree to within a few percent.