A statistical model is identifiable if the map parameterizing the model is injective. This means that the parameters producing a probability distribution in the model can be uniquely determined from the distribution itself which is a critical property for meaningful data analysis. In this talk I'll discuss a new strategy for proving that discrete parameters are identifiable that uses algebraic matroids associated to statistical models. This technique allows us to avoid elimination and is also parallelizable. I'll then discuss a new extension of this technique which utilizes oriented matroids to prove identifiability results that the original matroid technique is unable to obtain.
Vortragende: Sergio Albeverio, Niklas Boers, Lea Bossman, Jean Bricmont, Catalina Curceanu, Siddhant Das, Panel Discussion, Sandro Donadi, Fay Dowker, Rodolfo Figari, Jürg Fröhlich, Sheldon Goldstein, Michael Kiessling, Joel Lebowitz, Barry Loewer, Tim Maudlin, Franz Merkl, Sören Petrat, Alessandro Pizzo, Herbert Spohn, Session Short Talk, Stefan Teufel, Antoine Tilloy, Lev Vaidman, Howard Wiseman, Nino Zanghì
Link zur Info: https://laws-of-nature.net/conference2022/index.html
The skewproduct formalism is used to study some one-parametric bi- furcations in nonautonomous scalar ordinary differential equations. The concavity of the derivative of the right hand side precludes the existence of more than three minimal sets and ensures its hyperbolicity if there exist three. Thanks to these facts, saddle-node, transcritical and pitchfork bifurcations of minimal sets and discontinuous bifurcations of attractors are found. These bifurcations provide an adequate framework to study critical transitions: sudden and abrupt changes in the state of a complex system which occur on account of small variations on external parameters. This is joint work with Carmen Núñez and Rafael Obaya.