Approximately Bayesian Ensembling for Parameterized Neural ODEs

This repository contains the material related to the project of the Computational Statistics Course, Politecnico di Milano. The aim of the project was to explore Neural ODEs, implement methods to quantify predictions' uncertainty and implement Parameterized NODEs to model a 1-dimensional ODE-PDE.

Packages used:

• pytorch
• torchdiffeq