Entropy minimization and spectral dissimilarity curve resolution technique applied to nuclear magnetic resonance data sets.

The use of entropy minimization and spectral dissimilarity is applied to three nuclear magnetic resonance (NMR) data sets. The data sets contain 2, 2, and 3 observables each. It was found that without any a priori information the sets of pure component spectra underlying the NMR spectroscopic observations could be extracted. These successful spectral resolutions suggest that a combined entropy minimization and spectral dissimilarity approach can be further developed for even larger NMR data sets containing a larger number of observables. Brief comparison to DECRA and PMF curve resolution results is also presented.