AI Learns from the Past to Predict the Next Global Disaster
Dr. Zhiqin Ma's team trained machine learning models on surrogate copies of historical records, and in most tests these anticipated past tipping points better than standard warning indicators.












