A fast fine-grained genetic algorithm for spectrum fitting: An application to X-ray spectra

Often the result of the interaction of radiation with matter is represented by a spectrum. Regardless of
the type of radiation used, a spectrum consists of peaks superimposed on a background. The important
information is contained in the peak, so the derivation of the peaks parameters is fundamental. However,
in most cases the fitting of the spectrum is an ill-posed problem, because of the presence of superimposed
peaks. In this case the algorithms reported in the literature sometimes fail. For these reasons a genetic
algorithm approach has been developed and is described here. Its novelty for this kind of application
is represented by the use of a fine-grained strategy that makes it considerably faster than other
implementations of this class of algorithms. Some results of its application to X-ray spectra are presented
and discussed.