H. Bubert et H. Hillig, Artificial neural network and fuzzy clustering - New tools for evaluation of depth profile data?, MIKROCH ACT, 133(1-4), 2000, pp. 95-103
Depth profiling has been performed by using Auger electron spectrometry (AE
S) and X-ray photoelectron spectrometry (XPS) in combination with Ar-ion sp
uttering. The data obtained by both surface-analytical methods have been ev
aluated by means of factor analysis and afterwards by applying an artificia
l neural network or fuzzy clustering in order to determine the compositiona
l layering of different samples such as a Cr2O3/CrN sandwich layer, tarnish
layers on a nickel based alloy and on steel, and the coating of a Si3N4 ce
ramic powder. The applied artificial neural network was a Kohonen network.
It turned out that the method of fuzzy c-means clustering was more successf
ul than Kohonen network due to the fact that fuzzy c-means clustering start
s with more input information which can be obtained from factor analysis.