FPCA-Enhanced Simulation-Based Inference for Robust Type Ia Supernova Cosmology
概要
Precision cosmology is a cornerstone of modern astronomy, and upcoming wide-field photometric surveys make it crucial to develop robust, data-driven inference methods that can handle complex survey systematics without relying on closed-form likelihoods. Simulation-Based Inference (SBI) meets this need by enabling forward simulations that encode complex survey characteristics. Previous SBI analyses in supernova cosmology have used SALT2 light-curve parameters as summary statistics; however, SALT2…