Springer Complexity Lecture

Lisbon University Institute
13 September 2010, 6:00pm-7:00pm GMT, at the ECCS'10 Plenary Auditorium

Didier Sornette
"Parallels Between Earthquake Prediction, Financial Crash Prediction and Epileptic Seizures Predictions"


Abstract:
Earthquakes, financial crashes, and epileptic seizures are different catastrophic events, which are sometimes thought to reveal a common class of self-organized behavior found in many complex systems. The proposed analogy is often based on some common statistics, such as a power law distribution of event sizes, and on the existence of similar structures in terms of coupled threshold oscillators of relaxation. But, beyond the analogies, what do we learn and how does this help improve the state of the art in each relevant discipline? Here, we propose and document a correspondence between epileptic seizures (SZ) in humans and earthquakes (EQ) in Southern California, that extends quantitatively over seven distinct statistics, which exhibit remarkably robust scale-free power law properties. Then, based on a common theoretical framework, we test and verify the prediction that increasing the coupling strength and/or decreasing the heterogeneity should lead to the appearance of synchronized behavior with characteristic size and time scales. This is done for SZ by using rats treated intra-peritoneally with a convulsant. Going back from SZ to EQ, the proposed analogy informed by the positive tests on the rats suggests that characteristic earthquake behavior, previously proposed on the basis of controversial observations, should be a genuine class of dynamics in relevant seismo-tectonic settings characterized by strong seismic coupling and low crust heterogeneity. We also document similar patterns, coined "outliers'' or "kings'' in the distribution of financial drawdowns, which justify a resarch program for the prediction of crashes based on the hypothesis that they belong to a different class of their own and express specific amechanisms amplifying the normal volatility via positive feedbacks. We conclude with the potentials and state of the arts for the predictions of earthquakes, financial crashes and epileptic seizures.


Dr. Didier Sornette is Professor on the Chair of Entrepreneurial Risks at the Department of Management, Technology and Economics (D-MTEC), ETH Zurich; Director of the Financial Crisis Observatory; co-founder of the Competence Center for Coping with Crises in Socio-Economic Systems; member of the Swiss Finance Institute; Professor of Physics associated with the Department of Physics (D-PHYS), ETH Zurich; Professor of Geophysics associated with the Department of Earth Sciences (D-ERWD), ETH Zurich. His actual research priorities concern: (1) the Financial Crisis Observatory, a scientific platform aimed at testing and quantifying rigorously, in a systematic way and on a large scale the hypothesis that financial markets exhibit a degree of inefficiency and a potential for predictability, especially during regimes when bubbles develop (http://www.er.ethz.ch/fco/index); (2) the Center for the prediction of social, commercial and marketing success by combining information stemming from the dynamical responses to endogenous versus exogenous shocks of large databases; and (3) prediction of crises and extreme events in complex systems and risk management, with applications to social systems (financial crashes, recessions, cyber risks) and natural systems (earthquakes, rupture, epileptic seizures, immune system collapse).