We are pleased to announce the launch of an international coding competition on Kaggle. Participants will tackle a real-world transfer learning problem in spectroscopy: predicting the concentrations of glucose, acetate, and magnesium sulfate from Raman spectra — even across different instruments.
Why join?
-
Work on a cutting-edge machine learning challenge rooted in bioprocessing and analytical chemistry.
-
Access a unique dataset of more than 2,500 Raman spectra collected from eight different devices.
-
Test your model’s ability to generalize across hardware — a crucial hurdle in real-world Process Analytical Technology (PAT) applications.
-
Compete for one of three cash prizes: 1st – $750 | 2nd – $450 | 3rd – $300.
The competition is open now and will run for the next three months.
The challenge is hosted on Kaggle and supported by the EU-HORIZON DIG4BIO project together with KIWI-biolab. It is an opportunity to advance domain adaptation and contribute to the future of instrument-agnostic spectroscopy.
Learn more at: https://lnkd.in/dk-qjcrr