A synergy between reservoir computing and partial differential equations

What is it about?

PDEs can help reservoir to learn hidden features more accurately with a limited amount of data unlike other machine learning approaches such as deep learning for image denoising

Featured Image

Why is it important?

We have shown that classical PDE approaches can still play important roles in boosting the AI performances. While most machine learning algorithms work in a black box, classical PDE approaches imposing mathematical properties on the expected solution may tell the black box what and how to learn from the data.