Rapid optimization and minimal complexity in computational neural network multivariate calibration of chlorinated hydrocarbons using Raman spectroscopy
Wj. Egan et al., Rapid optimization and minimal complexity in computational neural network multivariate calibration of chlorinated hydrocarbons using Raman spectroscopy, J CHEMOMETR, 15(1), 2001, pp. 29-48
Improvements in the computational neural network modeling process are descr
ibed with the goals of enhancing the optimization process and reducing NN m
odel complexity. Improvements to the optimization process not only speed co
mputation, but also can enhance the quality of the result. Complex NN model
s require more intensive optimization procedures and are considerably more
difficult to interpret. Performance of these new algorithms is demonstrated
by results from training neural networks to quantitate composition of mixt
ures of chlorinated hydrocarbons based on their Raman spectra. Copyright (C
) 2000 John Wiley & Sons, Ltd.