AURORA from air pollution sources to mortality: modeling observed and latent exposures using machine learning

by Daniel Trejo Banos (SDSC/ETHZ)

Wednesday 15 May 2024, 11:30 → 13:00 Europe/Zurich

or https://psich.zoom.us/j/66556632390 (OHSA/E13)

ZOOM ID: https://psich.zoom.us/j/66556632390

Abstract:
Particulate Matter (PM) is a liquid or solid particle suspended in the air and is associated with around 8.9 million deaths worldwide in 2015. They are involved in many cardiovascular diseases, respiratory symptoms, cancer, diabetes, and neurodegenerative diseases. As part of this project of the 4th call, along with collaborators of the Paul Scherrer Institute and the Swiss Tropical and Public Health Institute, we have collected the most extensive observational dataset of air pollution in Europe, along with geographical and land use data for all the continent, accompanied by continent-wide simulations of air pollution dispersion for ten years. The final objective is to associate the pollutant concentrations to health outcomes using the Swiss National Cohort with health outcomes for 3 million individuals in Switzerland using survival analysis.

I will present a brief overview of the project, the challenges and machine learning approaches used for this project.