Hi,
I have spectral data ranging from 200-2600nm.For building AAN models, I have used PCA as data reduction method to reduce the input variables. Based on eigenvalues>1, 14 PC was chosen. Then using graph builder I have plotted the 14 PC vs the Wavelength (200-2600nm) to see which PC's contribute more in the separation and which wavelengths are more dominant. Now I have been questioned how this can be seen from PCA? Enclosed is the plot.
Thanks
