A Computer Program to Calculate the Percent Total Deuteration in an Exchange Reaction

The acid-catalyzed hydrogen-deuterium exchange reaction in aliphatic ketones was modeled as a purely random process. Using a computer program, a statistically random hydrogen-deuterium exchange was simulated for a particular molecule, 2-indanone. For this molecule, there are two carbons, each providing two potential sites for exchange. The program determined the relative amounts for each of the partially deuterated species, including the total percent deuteration. A comparison with mass spectral data showed good agreement with this model.     

PHOTOSENSITIZATION TISSUE MODEL

Photosensitization in a light scattering matrix was investigated with a tissue model consisting of polystyrene microsphere scattering particles, in the presence of a Photofrin II as a photosensitizing agent, and subtilisin Carlsberg as an enzyme target. The photodynamic rate constant for irradiation at 435 nm, 545 nm, and red light was measured at different microsphere concentrations. The reaction rate was almost independent of the microspheres, ruling out a significant effect of light scattering on the integrated photosensitization efficiency. Modeling with the one-dimensional diffusion approximation showed that increasing scatterer concentrations led to lower transmission and higher diffuse reflection, such that the fractional absorption was almost constant. The quantum efficiency of enzyme inactivation in the light-scattering systems was 0.0014 +/- 0.0003. In terms of incident dose, red and green light had approximately the same effectiveness and blue light was two-fold more efficient.     

A fraud-alert model for credit cards during the authorization process

^*Correspondence regarding this paper should be addressed to the first author In this paper, we set forth an expert-system model to help alertbanks and other financial institutions to fraudulent usage ofconsumer credit during the authorization process. The paperfirst addresses the model-building process, then briefly describesprototype development, and finally presents results from analysisof real data from a Canadian bank.     

Credit-scoring models for the evaluation of small-business loan applications

Received on 1 May 1991. Revision received 1 January 1991. In this paper, the decision process of commercial loan officersis modelled. Specifically, commercial loan applications froma major Canadian bank were analysed. Creditscoring models usingdiscriminant analysis and logit regression were created in aneffort to replicate the decision process. Results indicate thatcredit-scoring models for small-business loans are effectiveand that further analysis in this area is warranted.