Neural networks and extreme gradient boosting predict multiple thresholds and trajectories of microbial biodiversity responses due to browning
L. Fontaine, M. Khomich, T. Andersen, D. O. Hessen, S. Rasconi, M. L. Davey, A. Eiler
The paper uses machine learning to show non-monotonic relationships between microbial diversity and browning in freshwater lakes, revealing thresholds and alternative diversity trajectories.
bioRxiv · Mar 22, 2021 · PreprintMicrobiology