Speaker
Description
X-ray absorption spectroscopy provides a wealth of information regarding the local structure and electronic properties of materials. However, data analysis is significantly more time-consuming than acquisition and initial data reduction. Decoding the information relies on comparing it with similar compounds for which the spectrum–property mapping is already established, a task that is very often performed by visual inspection.
Machine learning (ML) is revolutionizing many fields with its ability to extract and learn patterns in big data without having to provide additional prior information other than the data itself. ML models give access to instantaneous predictions of properties and observables, which makes them particularly attractive for performing real-time analysis of the measured data or autonomous experimental acquisitions.
In this talk, I will present the different challenges faced when using machine learning models to analyze X-ray spectroscopic data, from building the initial training datasets to evaluating the ML models' robustness to the different sources of errors, such as spectral shift, normalization, noise level, and class imbalance, that limit the quality of the prediction.
| Abstract publication | I agree that the abstract will be published on the web site |
|---|
