Paper: AI for analysing World-Earth models

Deep Reinforcement Learning cannot only play Atari Games and Go but point to sustainable management pathways as well.
Paper: AI for analysing World-Earth models

PIK released this news item at the occasion of another bright master student's paper:

F.M. Strnad, W. Barfuss, J.F. Donges, J. Heitzig,
Deep reinforcement learning in World-Earth system models to discover sustainable management strategies,
Chaos 29, 123122 (2019),
DOI: 10.1063/1.5124673,
preprint arXiv:1908.05567 [physics.soc-ph] [cs.LG].

Happy holidays!