Speaker
Konstantinos Dialektopoulos
Description
In this talk, I will present a machine learning approach to reconstructing cosmological dynamics using artificial neural networks (ANNs), applied to late-time expansion, structure growth, and scalar-tensor gravity models. This non-parametric method learns directly from observational data—such as cosmic chronometers, BAO, and supernovae—while incorporating realistic uncertainties and correlations. I will highlight results from ΛCDM null tests and demonstrate how ANN-based reconstructions constrain viable scalar-tensor theories, offering a scalable and model-independent route to cosmological inference in the precision era.
