Transfer learning

The development of laser-induced breakdown spectroscopy (LIBS) reached the stars in terms of hardware; LIBS as a primary tool of SuperCam (Perseverance rover) landed on the planet Mars. On the contrary, the software development is hampered by the necessary ad hoc creation of data libraries for all LIBS systems with distinct analytical performances. This persists to be the main obstacle regardless of the LIBS application and it prevents further progress of data mining despite the boom of machine learning. Thus, the data libraries are the most valuable and irreplicable part of the LIBS system. Thus, we intend to establish the proof of concept for the transfer between distinct atmospheric conditions (e.g., Earth, Mars).