Dear colleagues and friends,

It is with great pleasure that we invite you to take part in a new data challenge for this edition of the EMSLIBS conference.

This challenge will focus on a classification problem. The goal is to predict the class of a sample from its LIBS spectrum acquired with a portable system. The spectral data provided for this challenge are raw signals directly obtained from the spectrometer.

All necessary data for the challenge are contained in a MATLAB file named data_challenge.mat, available for download here:
https://nextcloud.univ-lille.fr/index.php/s/8nENaMjXZLpszRJ

This MATLAB file contains the following matrices:

• spectra_cal (size: 582 rows × 23,431 columns): the 582 spectra to be used for training your predictive model. The spectral range consists of 23,431 wavelengths.
• classes_cal (size: 582 rows × 1 column): the class corresponding to each spectrum in spectra_cal. There are six classes in total, coded from 1 to 6.
• spectra_test_perm (size: 510 rows × 23,431 columns): the 510 spectra for which you must predict the associated classes.
• wave (size: 1 row × 23,431 columns): the wavelength values of the considered spectral range.

Challenge rules:

• The percentage of correct classifications estimated on the test set will serve as the figure of merit to rank the best approaches submitted by participants.
• Each participant must send a CSV file containing the predicted classes for the test set to Ludovic Duponchel by email (ludovic.duponchel@univ-lille.fr). At that time, you will be asked to provide your name and affiliation. If the work is carried out as a group, please include the list of all contributors.
• The submission deadline is January 16, 2026, at 12:00 (noon).
• Only registered conference attendees may participate in the challenge.
• Each participant may submit only one CSV file before the deadline. You should therefore select the approach that you consider most optimal.
• There are no restrictions on the tools or methodologies used.
• A dedicated challenge session will allow each of the top three participants to briefly present their approach. They will be contacted by the organizer during the week of January 19 to prepare a few slides for this presentation.
• To further highlight this initiative, the three best participants and the challenge organizers will jointly work on a publication in the months following the conference.

Please feel free to contact me for any questions regarding this challenge.

Good luck!
Ludovic Duponchel
Organizer of the EMSLIBS 2025 Challenge