B0717
Title: Uncertainty quantification in metric spaces
Authors: Marcos Matabuena - Harvard University (Spain) [presenting]
Gabor Lugosi - UPF (Spain)
Abstract: A novel uncertainty quantification framework is introduced for regression models where the response takes values in a separable metric space, and the predictors are in Euclidean space. The proposed algorithms can efficiently handle large datasets and are agnostic to the predictive base model used. Furthermore, the algorithms possess asymptotic consistency guarantees and, in some special homoscedastic cases, have non-asymptotic guarantees. To illustrate the effectiveness of the proposed uncertainty quantification framework, a linear regression model is used for metric responses (known as the global Frchet model) in various clinical applications related to precision and digital medicine. The different clinical outcomes analyzed are represented as complex statistical objects, including multivariate Euclidean data, Laplacian graphs, and probability distributions.
