Speaker
Description
Modern cosmology requires repeated evaluations of complex theoretical
predictions for inference, forecasts, and tests of physics beyond ΛCDM.
Symbolic regression provides a fast and accurate way to build surrogate
models while retaining compact analytical expressions that are
transparent, interpretable, and easy to implement.
I will present applications of symbolic regression to cosmological
observables, including CMB lensing, temperature, and polarization
spectra in extended ΛCDM models with massive neutrinos and evolving dark
energy. I will also discuss its use for matter power spectra in
non-standard dark matter scenarios, such as the Generalized Dark Matter
framework. These examples show that symbolic regression can be a
practical tool for accelerating cosmological analyses while keeping the
resulting models lightweight and analytically accessible.