Automated Machine Learning liberates domain scientists from selecting learners and hyperparameters and discovers the importance of atmospheric trace gases for improving surface PM2.5 estimates.
machine learning & AI
Posted inNews
“Icefin” Investigates a Glacial Underbelly
An instrument-laden submersible reveals where—and how rapidly—the Antarctic glacier is melting.
Posted inNews
The Limits to Tree Planting in the Indian Himalayas
The Indian government has an ambitious forestry goal. New research shows it may be out of sync with environmental and social constraints.
Posted inNews
Machine Learning Helps Researchers Track Illegal Fishing
Using machine learning, researchers found that nearly 20% of high seas fishing could be unauthorized.
Posted inResearch Spotlights
Prospecting for Copper with Machine Learning and Zircons
Using artificial intelligence, researchers can now identify zircons derived from valuable copper deposits.
Posted inNews
Boreal Trees May Grow Faster Due to Climate Change
Enhanced tree growth could significantly offset carbon emissions, but some researchers say it’s not enough to compete with forest disturbances.
Posted inNews
Mapping Rwanda’s Trees from Above
Researchers used both aerial and satellite imagery, as well as machine learning, to map the carbon stock of every overstory tree in Rwanda—the first such inventory in the world.
Posted inEditors' Highlights
Quantifying the Potential of Forestation for Carbon Storage
Forestation projects in southern China over the past few decades have sequestered large amounts of carbon in tree biomass, but the region is approaching saturation of forest carbon storage capacity.
Posted inEditors' Highlights
Playing Bricks with Neural Networks to Learn Sorption Processes
Designated neural network modules are combined to mimic numerically-discretized diffusion-sorption equations, which allows learning “missing pieces” in system understanding and their uncertainties.