Code and Data for Master's thesis: Certified Physics-Informed Neural Networks for Evolution Equations

This repository presents an implementation of certified Physics-Informed Neural Networks (PINNs) for solving partial differential equations with rigorous a posteriori error bounds. Building on the framework developed by Hillebrecht and Unger in 2022, the work extends the methodology to the Fisher–KPP equation. The dataset contains the results of five experimental studies, including the baseline replication and optimizer evaluation, as well as sensitivity analyses examining collocation density, architectural capacity, and temporal stability for the Fisher–KPP equation. The work was submitted as part of a Master’s thesis in 2026.