GENETIC PROGRAMMING - A NOVEL METHOD FOR THE QUANTITATIVE-ANALYSIS OFPYROLYSIS MASS-SPECTRAL DATA

A technique for the analysis of multivariate data by genetic programmi ng (GP) is described, with particular reference to the quantitative an alysis of orange juice adulteration data collected by pyrolysis mass s pectrometry (PyMS). The dimensionality of the input space was reduced by ranking variables according to product moment correlation or mutual information with the outputs. The GP technique as described gives pre dictive errors equivalent to, if not better than, more widespread meth ods such as partial least squares and artificial neural networks but a dditionally can provide a means for easing the interpretation of the c orrelation between input and output variables. The described applicati on demonstrates that by using the GP method for analyzing PyMS data th e adulteration of orange juice with 10% sucrose solution can be quanti fied reliably over a 0-20% range with an RMS error in the estimate of similar to 1%.