STAT-IR 7

The SMIS beamline at SOLEIL synchrotron is pleased to announce the 7th edition of its school, STAT-IR 7, which is dedicated to training scientists in the field of multivariate and machine learning analysis of infrared microspectroscopy data. The course is scheduled to take place from Monday 16th to Wednesday 18th October 2023 at the SOLEIL synchrotron and will be conducted in English.
Computers will be available in the training room (personal computers are accepted).

The training is limited to 20 participants.

The aim of this school is to provide attendees with hands-on training in the use of Multivariate Analysis (MVA) techniques like PCA, and Machine Learning (ML) methods such as unsupervised and supervised clustering for analyzing infrared microspectroscopy data. The course will also cover spectral preprocessing techniques used in preparation for MVA and ML. Additionally, the applications of MVA and ML techniques for analyzing hyperspectral images and maps will be taught. The training program will emphasize the use of biomedical datasets, though the methodologies learned can also be applied to data related to cultural heritage and polymer sciences. The techniques can also be applied to data obtained from Raman microspectroscopy.

Infrared microspectroscopy and imaging provide essential insights into the chemical composition, spatial organization, and molecular conformation of complex materials with spatial resolutions ranging from microns to submicron. Data analysis plays a pivotal role in unravelling the information-rich content within the infrared spectra of complex systems and numerous specific techniques can be employed. With the use of modern instrumentation large datasets comprising tens to hundreds of thousands of spectra can be rapidly collected. Therefore, it is necessary to use automated techniques to extract information.

One of the most successful approaches to exploit large datasets lies on statistical analyses to explore their spectral variability, find correlations between data, perform objective classifications, train identification algorithms, correlate data from different techniques, or establish predictive models for quantification. Such methods can only be successful if specific spectral preprocessing methods are applied to remove artefacts and confusing variations. Specific software, preprocessing techniques and strategies must be applied requiring specialized training combining signal processing, MVA and ML.

The SMIS beamline is the infrared microspectroscopy beamline of SOLEIL synchrotron facility located in Gif-sur-Yvette near Paris. The SMIS beamline staff has 15-year experience in the measurement and analysis of microspectroscopy data. The course will be given by Ferenc Borondics (expert in physical chemistry), Marko Toplak (machine learning specialist and one of Quasar developers), and Christophe Sandt (biospectroscopist).

The course will be done with the open-source software Quasar. Hand-on session with participant data will be held.

The formation will be held on site at SOLEIL synchrotron and opened to all past and future users of the beamline.

The cost for the formation will be 66€ TTC.

Registration is closed (limit of seats reached).

For further information please contact: Conf-statir7-2023@synchrotron-soleil.fr