Dr. Sergey Guda
(Southern Federal University)
Quantitative analysis of the XANES spectra with elements of machine learning
26/10/2018 Physical department room 219
Participant limit 25 persons
Preliminary registration is needed.
This course will provide a short theoretical introduction to machine learning methods for applications in spectroscopy. Such problems as learning datasets, cross-validation, overtraining and undertraining will be revised. Neural networks for the deep learning will be briefly discusses. Practical session will cover application of the PyFitIt software to the quantitative analysis of XANES spectra. Attendees wil learn how to generate structures for training sample, calculate spectra on Blokhin supercomputer, perform approximation over difference spectra, apply direct and indirect approaches to the refinement problem. In the end of the course student will learn how to apply machine learning methods to any kind of spectroscopy data.