A Comparison of some Computer-measured and Human-measured Pattern Recognition Properties

Abstract

Some properties which have been useful in the computerized classification of white blood cells and breast tumors are compared to similar properties measured by human beings. It is shown that the computer properties need not be close approximations to the human properties for them to be valuable.     

The Fusion Science Research Plan for the Major U.S. Tokamaks

This is the May 1996 report of a subpanel of the US Department of Energy Fusion Energy Sciences Advisory Committee (FESAC), charged with conducting a review of the progress, priorities and potential near-term contributions of TFTR, DIII-D and Alcator C-MOD (and other facilities as appropriate) as part of the transition to a Fusion Energy Sciences Program and produce an optimum plan for obtaining the most scientific benefit from them.     

Unit to unit variation of pesticide residues options for acute dietary risk assessment

Within the context of the current UK policy for those commodities where the potential variability of residues in individual commodity units is taken into account, a variability factor nu , which reflects the ratio of a high level residue in the individual commodity unit to the corresponding composite residue level, is used. Data gathered from supervised industry trials in which individual items were analysed following treatment, although limited, show that variability is typically lower than that reflected by the default factors currently used and that the range of variability is reproducible over the limited range of different a.i./crop/method of application combinations investigated. In order to improve the accuracy of the acute dietary exposure estimate, the European Crop Protection Association (ECPA) proposes the following alternative to the current Tier I approach. The residue level input from a 'hot' unit within the dietary risk assessment should be determined using the highest composite sample residue from supervised field trials and a generic variability factor (nu) determined experimentally from supervised trials. The variability factor itself should be calculated as the 95th percentile level of the residue level found in an individual unit (or single serving portion for large crops) divided by sample mean for data produced from supervised trials. This would improve the accuracy of the Tier I approach and allow attention to be focused on particular a.i./crop/method of application combinations where the NESTI> acute RfD (based on a Tier I assessment) and generation of individual unit residue data for the particular outlet or other mitigation may be appropriate.     

Deactivation of ICI low temperature methanol catalyst in an industrial reactor

The deactivation behavior of one of the ICI low temperature methanol catalysts (ICI-51-Z) in a typical ICI multibed quench reactor has been studied in relation to methanol production with respect to time of operation, using some representative sets of plant data collected over two years time span. A pseudo-homogeneous first order model of the methanol synthesis reactor was formulated for this purpose. The best set of catalyst deactivation parameters for different beds was found using a nonlinear parametric estimation technique. Catalyst in the entering bed was found to deactivate most with the time of operation, while subsequent beds deactivate progressively less quickly.     

Fractionation of soybean phospholipids by high-performance liquid chromatography with an evaporative light-scattering detector

Phosphatidylcholine (PC) and phosphatidylethanolamine (PE) from 23 soybean lines with a wide range of fatty acid compositions were resolved into seven fractions by high-performance liquid chromatography (HPLC). Fraction identities were assigned from fatty acid compositions determined by gas chromatography (GC). A mass detector, i.e., an evaporative light-scattering detector, was used for HPLC quantification. The detector response was a power function of PC and PE concentrations. Various correction methods were applied to the detector response to obtain the best agreement between phospholipid (PL) fatty acid compositions determined by GC and that calculated from the corrected HPLC fraction percentages. The corrected HPLC fraction composition also was compared with that calculated from stereospecific distribution data using a 1-random-2-random hypothesis. Correlation between PL-fatty acid and HPLC-fraction percentages showed that genetic modification of soybean oil composition caused changes in PL species, which alter physical properties and may alter the physiological functions of PL in biomembranes.     

Influence of Polydimethylsiloxane on the Oxygen Concentration of Oils at Various Temperatures

The effect of temperature on the oil oxygen concentration, tested in both soybean and olive oils with no added polydimethylsiloxane (PDMS), showed that the oxygen concentration increased with temperature to approximately 100 °C. Above 100 °C, the oxygen concentration abruptly decreased. This change was attributed to the balance between the rates of oxygen uptake and consumption by oil oxidation, which favored oxygen consumption over uptake at temperatures above 100 °C. The addition of 100 ppb PDMS to soybean oil, enough to form a continuous layer over the surface of the oil, reduced the oxygen concentration when compared to a soybean oil control containing no added PDMS at temperatures ranging from 93 to 180 °C; thus suggesting an oxygen barrier effect of PDMS. The accumulation of PDMS at the air–oil interface in soybean oil held at 180 °C was determined by comparing the oil’s internal temperature and the apparent surface temperature. A decrease in the apparent surface temperature while the oil was held at a constant internal temperature was attributed to a change in the emissivity of the surface as a consequence of the accumulation of PDMS in the air–oil interface. The presence of PDMS at the air–oil interface was confirmed for 100 ppm of PDMS, a concentration greater than the concentration necessary to form a monolayer of PDMS on the oil surface.