Key points are not available for this paper at this time.
Abstract : The purpose of the research is the development of mathematical methods and computer programs for the extraction of meaningful information from biological, primarily neurophysiological, measurements. Emphasis was placed on statistical methods suitable for separating two or more random signals and which provide insight into the underlying mechanism by which the signals are generated. Loeve-Karhunen expansion and discriminant analysis methods are applied to the problem of time signal classification. Experiments are performed both on computer generated time signals and on electroencephalograms. Methods of coping with the singularity problem arising from a small sample size are investigated.
Streeter et al. (Fri,) studied this question.