Automated assessment of vigilance using committees of radial basis function analysers

The analysis of human vigilance (in terms of the level of subject alertness) on the basis of a small number of physiological measures is considered, Results from an initial feature selection phase are shown and the prediction of the human scoring process using committees of radial basis function (RBF) classifiers is considered. Comparative results are shown for regression-based and classification-based analysis on a sample of representative data sets