Some aspects of the analysis of coal by X-ray fluorescence spectroscopy

Bituminous coal samples from 84 distinct sources were analysed by conventional British Standards (BS) methods for phosphorus, sulphur, chlorine, ash and the ash-forming elements (calcium, silicon, aluminium, iron, sodium, magnesium, potassium, titanium and manganese). In most cases four determinations were made per element per source. Samples were crushed to ?212 μm (72 BS mesh) and pressed into aluminium cups at 20 t in^?2 (309 MPa) without binder or backing. Two pellets from each sample were analysed by wave-length dispersive X-ray fluorescence techniques, using a spectrometer equipped with a rhodium rube; the X-ray count took 60 s. The X-ray results were calibrated against the conventional results using multiple regression. The accuracy obtained was comparable with routine ‘wet chemical’ techniques. The X-ray technique is suitable for the routine determination of large numbers of samples.     

Isolation and characterization of ganglioside 9-O-acetyl-GD3 from bovine buttermilk

Bovine buttermilk contains a unique ganglioside, 9-O-acetyl-GD₃. In order to isolate large quantities of this ganglioside, a simplified isolation scheme which consists of several ion-exchange and silica gel column chromatographic procedures was devised. The isolated 9-O-acetyl-GD₃ was characterized on the basis of its thin-layer chromatographic behavior, its immunoreactivity with a specific monoclonal antibody, JONES, and by conversion to authentic GD₃ by mild base treatment.     

Colin Allen and Marc Bekoff,Species of Mind: The Philosophy and Biology of Cognitive Ethology

Minds and Machines -     

A multidomain spectral method for solving elliptic equations

We present a new solver for coupled nonlinear elliptic partial differential equations (PDEs). The solver is based on pseudo-spectral collocation with domain decomposition and can handle one- to three-dimensional problems. It has three distinct features. First, the combined problem of solving the PDE, satisfying the boundary conditions, and matching between different subdomains is cast into one set of equations readily accessible to standard linear and nonlinear solvers. Second, touching as well as overlapping subdomains are supported; both rectangular blocks with Chebyshev basis functions as well as spherical shells with an expansion in spherical harmonics are implemented. Third, the code is very flexible: The domain decomposition as well as the distribution of collocation points in each domain can be chosen at run time, and the solver is easily adaptable to new PDEs. The code has been used to solve the equations of the initial value problem of general relativity and should be useful in many other problems. We compare the new method to finite difference codes and find it superior in both runtime and accuracy, at least for the smooth problems considered here.     

On the distribution of performance from multiple neural-networktrials

The performance of neural network simulations is often reported in terms of the mean and standard deviation of a number of simulations performed with different starting conditions. However, in many cases, the distribution of the individual results does not approximate a Gaussian distribution, may not be symmetric, and may be multimodal. We present the distribution of results for practical problems and show that assuming Gaussian distributions can significantly affect the interpretation of results, especially those of comparison studies. For a controlled task which we consider, we find that the distribution of performance is skewed toward better performance for smoother target functions and skewed toward worse performance for more complex target functions. We propose new guidelines for reporting performance which provide more information about the actual distribution.     

Polarization Phase-Based Method For Material Classification In Computer Vision

A robust and accurate polarization phase-based technique for material classification is presented. The novelty of this technique is three-fold in (i) its theoretical development, (ii) application, and, (iii) experimental implementation. The concept of phase of polarization of a light wave is introduced to computer vision for discrimination between materials according to their intrinsic electrical conductivity, such as distinguishing conducting metals, and poorly conducting dielectrics. Previous work has used intensity, color and polarization component ratios. This new method is based on the physical principle that metals retard orthogonal components of light upon reflection while dielectrics do not. This method has significant complementary advantages with respect to existing techniques, is computationally efficient, and can be easily implemented with existing imaging technology. Experiments for real circuit board inspection, nonconductive and conductive glass, and, outdoor object recognition have been performed to demonstrate its accuracy and potential capabilities.     

Sindi: an intelligent assistant for highway design

We present a system to support highway engineers in the tasks of preliminary highway design. The system is based on the ideas of heuristic search and explicit preference setting. The system has been implemented and installed in a Spanish engineering firm.