Emily Chan
“For example, in a financial system, a node could be a single company; in an ecological system, a node could denote a user, and so forth,” Yan said.
Then, the researchers trained their model by feeding it data from simple theoretical systems. Afterward, they finally gave the model a real-world problem to analyze—the transformation of tropical forests into savannah.
They took more than 20 years of data on rainfall and tree coverage in three regions in Central Africa and fed the algorithm information on two of those regions.
The AI was able to accurately predict what happened in the third region. Now, the researchers hope to apply the model to other systems like pandemics, financial crashes and wildfires. They will focus on parts of human systems that cannot be affected by human intentions.
For instance, when drivers are aware of which roads are congested, they may alter their routes. Some roads may become less crowded, but others may get worse. It’s a challenge that comes with predicting systems.
“Using AI to capture these fundamental signals can be valuable for making predictions,” Yan said.
“Although predicting such systems is challenging, it is worthwhile because critical transitions in human-involved systems can have even more severe consequences.”
The study was published in the journal Physical Review X.
