Discrimination of Venezuelan spirituous beverages by a trace element-radial basis neural network approach.

Radial basis neural networks (RBNNs) were developed and evaluated for discrimination of specimens of 'aguardiente de Cocuy', a spirituous beverage produced in the northwestern region of Venezuela. The beverage is distilled from the must of Agave cocui Trelease in an artisanship fashion with little quality control. Forty specimens, with known concentrations of copper, iron, and zinc, were used in this study. The specimens were previously collected in various locations around Sucre Municipality (Falcón State) and Urdaneta Municipality (Lara State). The normalized concentrations of these elements served as indirect descriptors of origin (input data). They were presented to the neural networks through 1-3 input nodes in seven different combinations. In addition, two categories (two collection sites) and four categories (two collection sites+two manufacturing conditions) were designated as output data, in order to assess the impact of such selection on the discrimination performance. The overall performance of the four-category RBNNs was as follows (the input data is indicated in parentheses): (Cu-Fe)>(Cu-Zn)>(Cu)>(Zn)>(Fe-Zn)>(Cu-Fe-Zn)>(Fe). In this case, the highest percentage of correct hits was 82.5%. For the two-category RBNNs, the performance decreased as indicated below: (Cu)>(Cu-Fe)>(Cu-Zn)>(Fe-Zn)>(Zn) approximately (Cu-Fe-Zn)>(Fe). The reduction in the number of categories led to an increase in the discrimination performance of all the RBNNs, the best of which was 90.0%. The possibility of discriminating specimens of 'aguardiente de Cocuy' with such an accuracy, based on a single-element determination, is particularly attractive as it would result in a reduction of analysis' costs and laboratory's response time.

Evaluation of a generalized regression artificial neural network for extending cadmium's working calibration range in graphite furnace atomic absorption spectrometry.

A generalized regression artificial neural network (GRANN) was developed and evaluated for modeling cadmium's nonlinear calibration curve in order to extend its upper concentration limit from 4.0 microg L-1 up to 22.0 microg L-1. This type of neural network presents important advantages over the more popular backpropagation counterpart which are worth exploiting in analytical applications, namely, (1) a smaller number of variables have to be optimized, with the subsequent reduction in "development hassle"; and, (2) shorter development times, thanks to the fact that the adjustment of the weights (the artificial synapses) is a non-iterative, one-pass process. A backpropagation artificial neural network (BPANN), a second-order polynomial, and some less frequently employed polynomial and exponential functions (e.g., Gaussian, Lorentzian, and Boltzmann), were also evaluated for comparison purposes. The quality of the fit of the various models, assessed by calculating the root mean square of the percentage deviations, was as follows: GRANN>Boltzmann>second-order polynomial>BPANN>Gauss>Lorentz. The accuracy and precision of the models were further estimated through the determination of cadmium in the certified reference material "Trace Metals in Drinking Water" (High Purity Standards, Lot No. 490915), which has a cadmium certified concentration (12.00+/-0.06 microg L-1) that lies in the nonlinear regime of the calibration curve. Only the models generated by the GRANN and BPANN accurately predicted the concentrations of a series of solutions, prepared by serial dilution of the CRM, with cadmium concentrations below and above the maximum linear calibration limit (4.0 microg L-1). Extension of the working range by using the proposed methodology represents an attractive alternative from the analytical point of view, since it results in less specimen manipulation and consequently reduced contamination risks without compromising either the accuracy or the precision of the analyses. The implementation of artificial neural networks also helps to reduce the trial-and-error task of looking for the right mathematical model from among the many possibilities currently available in the various scientific and statistic software packages.