Neural ODE solvers

Presentation and notebook with four assignments on neural ODE solvers.

The materials were designed for AI Tech spring school held in May 2022 at Politechnika Gdańska, Poland.

Presentation

Slides from the presentation are available in the neural_ode_solvers_presentation.pdf file. They are divided into sections:

1. Introduction.
2. Numerical ODE solvers.
3. Neural network as an ODE (covering the adjoint method).
4. Continuous Normalizing Flow (CNF).

Notebook

The neural_ode_solvers.ipynb file contains solved exercises with descriptions, instructions, and unit tests. You can find four models on the following topics:

1. Linear system of two ODEs - stationary dynamic function.

2. 1D ODE - non-stationary dynamic function.

3. MNIST classification. Classifier achieves ~97% accuracy.

4. Continuous Normalizing Flow (CNF).