Events

Date: 20 November 2025   Time: 14:00 - 15:00

Location: Hybrid: Seminar Room MB-503, School of Mathematical Sciences, QMUL, or via the Teams link below

In the talk I will address the growing concern of cascading extreme events, such as an extreme earthquake followed by a tsunami, by presenting a novel risk assessment method for these domino effects. The proposed approach develops an extreme value theory framework within a Kolmogorov–Arnold network (KAN) to estimate the probability of one extreme event triggering another, conditionally on a feature vector. An extra layer is added to the KAN's architecture to enforce the definition of the parameter of interest within the unit interval, and we refer to the resulting model as KANE (KAN with Natural Enforcement). The proposed method is backed by exhaustive numerical studies and further illustrated with real-world applications. Finally, I will provide some hints on how the framework of cascading extremes offers natural links with generative AI modeling of extreme events.

Short Bio: Miguel de Carvalho is the Chair of Statistical Data Science at the University of Edinburgh (UoE) and Honorary Professor at Universidade de Aveiro. He is elected fellow of the Generative AI Lab (UoE), co-director of the Edinburgh Centre for Financial Innovations, member of the Council of the International Statistical Institute, and past member of the board of directors of the International Society for Bayesian Analysis. Miguel's research interests include, inter alia, extreme value theory, Bayesian analysis, and the interfaces between Statistics and AI. He has been an AE for a variety of top tier journals such as Bayesian Analysis, The American Statistician, The Annals of Applied Statistics, and the Journal of the American Statistical Association. Miguel co-chaired the international conference EVA 2021 in Edinburgh and co-edited the Extremes special issue Bridging Heavy Tails & AI.

Teams link