Predicting critical transitions wit被国际著名加拿大科学评价追踪网站Science Featured Series以 “AI Learns from the Past to Predict the Next Global Disaster为题对该工作专题报道。edicting critical transitions withPredicting critical transitions withmachine learning trained on surrogatesof historical datamachine learning trained on surrogatesZhiqin Ma 1, Chunhua Zeng 1 , Yi-Cheng Zhang2 & Thomas M. Bury 3

Critical transitions can occur in many natural and man-made systems. Generic early warning signalsCritical transitions can occur in many natural and man-made systems. Generic early warning signalsCritical transitions can occur in many natural and man-made systems. Generic early warning signals被国际著名加拿大科学评价追踪网站Science Featured Series以 “AI Learns from the Past to Predict the Next Global Disaster为题对该工作专题报道。motivated by dynamical systems theory have had mixed success on real noisy data. More recentstudies found that deep learning classifiers trained on synthetic data could improve performance.However, to the best of our knowledge, neither of these methods take advantage of historical, system-specific data. Here, we introduce an approach that trains machine learning classifiers on surrogatedata of past transitions. The approach provides early warning signals in empirical and experimentaldata with higher sensitivity and specificity than two widely used generic early warning signals—variance and lag-1 autocorrelation. Since the approach is trained on surrogates of historical data, it isnot bound by the restricting assumption of a local bifurcation like previous methods. This system-specific approach can contribute to improved early warning signals to help humans better prepare foror avoid undesirable critical transitions.motivated by dynamical systems theory have had mixed success on real noisy data. More recentstudies found that deep learning classifiers trained on synthetic data could improve performance.However, to the best of our knowledge, neither of these methods take advantage of historical, system-specific data. Here, we introduce an approach that trains machine learning classifiers on surrogatedata of past transitions. The approach provides early warning signals in empirical and experimentaldata with higher sensitivity and specificity than two widely used generic early warning signals—variance and lag-1 autocorrelation. Since the approach is trained on surrogates of historical data, it isnot bound by the restricting assumption of a local bifurcation like previous methods. This system-specific approach can contribute to improved early warning signals to help humans better prepare foror avoid undesirable critical transitions.motivated by dynamical systems theory have had mixed success on real noisy data. More recentstudies found that deep learning classifiers trained on synthetic data could improve performance.However, to the best of our knowledge, neither of these methods take advantage of historical, system-specific data. Here, we introduce an approach that trains machine learning classifiers on surrogatedata of past transitions. The approach provides early warning signals in empirical and experimentaldata with higher sensitivity and specificity than two widely used generic early warning signals—variance and lag-1 autocorrelation. Since the approach is trained on surrogates of historical data, it isnot bound by the restricting assumption of a local bifurcation like previous methods. This system-specific approach can contribute to improved early warning signals to help humans better prepare foror avoid undesirable critical transitions.of historical dataCheck for updatesZhiqin Ma 1, Chunhua Zeng 1 , Yi-Cheng Zhang2 & Thomas M. Bury 3Critical transitions can occur in many natural and man-made systems. Generic early warning signalsmotivated by dynamical systems theory have had mixed success on real noisy data. More recentstudies found that deep learning classifiers trained on synthetic data could improve performance.However, to the best of our knowledge, neither of these methods take advantage of historical, system-specific data. Here, we introduce an approach that trains machine learning classifiers on surrogatedata of past transitions. The approach provides early warning signals in empirical and experimentaldata with higher sensitivity and specificity than two widely used generic early warning signals—variance and lag-1 autocorrelation. Since the approach is trained on surrogates of historical data, it isnot bound by the restricting assumption of a local bifurcation like previous methods. This system-specific approach can contribute to improved early warning signals to help humans better prepare foror avoid undesirable critical transitions.machine learning trained on surrogatesof historical dataCheck for updatesZhiqin Ma 1, Chunhua Zeng 1 , Yi-Cheng Zhang2 & Thomas M. Bury 3Critical transitions can occur in many natural and man-made systems. Generic early warning signalsmotivated by dynamical systems theory have had mixed success on real noisy data. More recentstudies found that deep learning classifiers trained on synthetic data could improve performance.However, to the best of our knowledge, neither of these methods take advantage of historical, system-specific data. Here, we introduce an approach that trains machine learning classifiers on surrogatedata of past transitions. The approach provides early warning signals in empirical and experimentaldata with higher sensitivity and specificity than two widely used generic early warning signals—variance and lag-1 autocorrelation. Since the approach is trained on surrogates of historical data, it isnot bound by the restricting assumption of a local bifurcation like previous methods. This system-specific approach can contribute to improved early warning signals to help humans better prepare foror avoid undesirable critical transitions.