Applications of self-organizing maps (SOM) to the composition determination of chemical products

Self-organizing maps (SOM) method was developed by Kohonen (1997) was first applied to the problems of chemical analysis, such as AES (Auger electron spectroscopy), XPS (X-ray photoelectron spectroscopy), and XRD (X-ray diffraction) data. Using the constructed 2-dimensional SOM, it became clear that the things which are described quantitatively by linguistic expressions can be explained more quantitatively by the position of the spectra data on the SOM as well as with a gray level expression. Also, the composition of the unknown spectra can be determined very precisely by the SOM, which is constructed using the composition of the known spectra.