Weak regularization by degenerate Lévy noise and its applications
Abstract: After a general introduction about the regularization by noise phenomenon in the degenerate setting, the first part of this PhD thesis focuses at establishing the Schauder estimates, a useful analytical tool to prove also the well-posedness of stochastic differential equations (SDEs), for two different classes of Kolmogorov equations under a weak H\"ormander-like condition, whose coefficients lie in suitable anisotropic H\"older spaces with multi-indices of regularity. The first class considers a nonlinear system controlled by a symmetric stable operator acting only on some components. Our method of proof relies on a perturbative approach based on forward parametrix expansions through Duhamel-type formulas. Due to the low regularizing properties given by the degenerate setting, we also exploit some controls on Besov norms, in order to deal with the non-linear perturbation. As an extension of the first one, we also present Schauder estimates associated with a degenerate Ornstein-Uhlenbeck operator driven by a larger class of stable-like operators, like the relativistic or the Lamperti stable one. Exploiting a backward parametrix approach, the second part of this work aims at establishing the weak well-posedness for a degenerate chain of SDEs driven by the same class of stable-like processes, under the assumptions of the minimal H\"older regularity on the coefficients. As a by-product of our method, we also present Krylov-type estimates of independent interest for the associated canonical process. Finally, we show through suitable counter-examples the existence of an (almost) sharp threshold on the regularity exponents ensuring the weak well-posedness for the SDE. In connection with some possible applications to kinetic dynamics with friction, we conclude by investigating the stability of second-order perturbations for degenerate Kolmogorov operators in Lp and H\"older norms.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.