CCSS Meeting #58: An introduction to scientific modelling with neural ODEs

This lecture will be held in hybrid format: the speaker Dr. Patrick Kidger will be speaking online from US, and Prof. dr. Henk Dijkstra will be physically present as the on-site moderator to open the session and initiate discussions. The theme of this CCSS Lunch Meeting is Neural Differential Equations and the applications in Complex Systems.

In the CCSS living room, participants can enjoy refreshments before and dinner afterwards - please signup for free below.

Speaker Overview

Dr. Patrick Kidger is a mathematician at Google X. He previously completed my PhD at the Mathematical Institute of the University of Oxford, studying machine learning, and in particular neural differential equations. In his work, he is probably best known for two things:

Lecture Overview

This talk will offer a first tutorial on neural ODEs for scientific applications. These combine many of the advantages of both differential equations and deep learning -- arguably the two most successful modelling paradigms ever developed. The goal of this introduction is to (a) provide a powerful modelling tool that complements existing theory-driven approaches, (b) demonstrate that they are easy to use via modern autodifferentiable software, and (c) give enough of the tips-and-tricks needed to make them work in practice!

Meeting Details

16:30-16:45 Walk in and refreshments
16:45-17:00 General introduction to use of machine learning in complex systems by Prof. dr. Henk Dijkstra (onsite)
17:00-17:30 An introduction to scientific modelling with neural ODEs by Dr. Patrick Kidger (online)
17:30-17:45 Q&A
17:45-18:30 Networking & Dinner (Pizza and/or Pasta!)

To attend the lecture online, please click the Zoom link at 16:45 on Tuesday 19 September.

To attend the lecture & the dinner (physically), please signup below before 15:00 on Monday 18 September.