Gabriele Visentin


Last Name

Visentin

First Name

Gabriele

ORCID

0000-0002-5198-6856

Organisational unit

02204 - RiskLab / RiskLab

Filters

2024 - 20252
Reset filters

Search Results

Publications 1 - 2 of 2
  • Computing Optimal Transport Maps and Wasserstein Barycenters Using Conditional Normalizing Flows
    Item type: Conference Paper
    Visentin, Gabriele; Cheridito, Patrick (2025)
    Proceedings of Machine Learning Research ~ Proceedings of the 42nd International Conference on Machine Learning
    We present a novel method for efficiently computing optimal transport maps and Wasserstein barycenters in high-dimensional spaces. Our approach uses conditional normalizing flows to approximate the input distributions as invertible pushforward transformations from a common latent space. This makes it possible to directly solve the primal problem using gradient-based minimization of the transport cost, unlike previous methods that rely on dual formulations and complex adversarial optimization. We show how this approach can be extended to compute Wasserstein barycenters by solving a conditional variance minimization problem. A key advantage of our conditional architecture is that it enables the computation of barycenters for hundreds of input distributions, which was computationally infeasible with previous methods. Our numerical experiments illustrate that our approach yields accurate results across various high-dimensional tasks and compares favorably with previous state-of-the-art methods.
  • A default system with overspilling contagion
    Item type: Journal Article
    Coculescu, Delia; Visentin, Gabriele (2024)
    Frontiers of Mathematical Finance
    Some dynamical contagion models for default risk have been pro-posed in the literature, where a system (composed of individual debtors) evolves as a Markov process conditionally on the observation of its stochastic environ-ment, with interacting intensities. The Markovian assumption necessitates that the environment evolves au-tonomously and is not influenced by the transitions of the system. We extend this classical literature and allow a default system to have a contagious impact on its environment. With a certain probability, the transition of a debtor to the default state has an impact on the system’s environment. This in turn affects the transition intensities of the other debtors inside the system. Therefore, in our framework, contagion can either be contained within the default system (i.e., direct contagion from a counterparty to another) or spill from the default system over its environment (indirect contagion). This type of model is of interest whenever one wants to capture within a model possible impacts of the defaults of a class of debtors on the more global economy and vice versa.
Publications 1 - 2 of 2