Project 11

Machine learning techniques in X-ray spectroscopy

The Terry lab has been developing machine learning software for the analysis of x-ray spectroscopy data. The two images on the left show spectra fit using a machine learning algorithm. The left image is a C 1s photoelectron spectroscopy spectrum from a polymer sample. The middle image shows a fit X-ray absorption fine structure spectrum from a metal alloy that was fit using the same basic algorithm. The advantage of this particular type of machine learning algorithm is that the same code base can be applied to multiple types of spectroscopies. The next step in the development is to be able to search databases for parameters that can be used to populate the initial fits in the machine learning code.