Keyan Miao (Interfacing Control with Neural ODEs)
From Coffee and Control · · 1 hr 0 min
This plays from the publisher's own server. We do not host a copy.
Description
In this episode I got to chat with the incredibly wonderful Dr Keyan Miao about the fascinating work she’s done on neural ODEs. We discuss applying control theoretic tools to optimise neural ODEs, as well as their use for both controller and observer design. And in case that wasn’t already enough we also touch on dissipativity, Koopman operators, and reinforcement learning!
You can find the transcript for this episode at: https://docs.google.com/document/d/1iI0swQSnNEqgGQgf7og8udRV3IBCXbbfEkLFlolGxEU/edit?usp=sharing
Follow the podcast on LinkedIn: https://www.linkedin.com/company/coffee-and-control-podcast/
Or connect with me directly: https://www.linkedin.com/in/lucy-hodgins-733a30175/
Join the live recording on 20th October 2026!
https://www.youtube.com/watch?v=2S9X66_Wfzc
( 9:30 PDT, 12:30 ET, 16:30 UTC, 17:30 BST, 18:30 CEST)
References
Keyan Miao: https://kymiao.github.io/
Richard Vinter: https://profiles.imperial.ac.uk/r.vinter
QianXiao Li from NUS and Wei-Nan Er from Princeton: https://www.jmlr.org/papers/volume18/17-653/17-653.pdf (https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.jmlr.org%2Fpapers%2Fvolume18%2F17-653%2F17-653.pdf&data=05%7C02%7CL.Hodgins%40soton.ac.uk%7Ceffa94a12d09446b9aac08dedac7f44e%7C4a5378f929f44d3ebe89669d03ada9d8%7C0%7C0%7C639188749665599257%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=i7GMI92aFZ%2Bns5pIh%2BdNRcceZv5qQH9csXgyoS%2BL988%3D&reserved=0)
https://link.springer.com/article/10.1007/s40687-018-0172-y (https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Flink.springer.com%2Farticle%2F10.1007%2Fs40687-018-0172-y&data=05%7C02%7CL.Hodgins%40soton.ac.uk%7Ceffa94a12d09446b9aac08dedac7f44e%7C4a5378f929f44d3ebe89669d03ada9d8%7C0%7C0%7C639188749665617855%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=xWFhX%2FrJKuig%2BeSkqd5v%2BYaBLAHWMjNIRQX6ebFfjQM%3D&reserved=0)
https://ems.press/journals/jems/articles/5404458 (https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fems.press%2Fjournals%2Fjems%2Farticles%2F5404458&data=05%7C02%7CL.Hodgins%40soton.ac.uk%7Ceffa94a12d09446b9aac08dedac7f44e%7C4a5378f929f44d3ebe89669d03ada9d8%7C0%7C0%7C639188749665627348%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=enqWf9hUa3OOVcB9VnbA1176ye6a24U1rsMiB40MdSo%3D&reserved=0)
Book on control theory/ philosophy: https://apac.ee.kntu.ac.ir/publications/books/control-systems-a-historical-and-philosophical-perspective/
Control theory/ philosophy article: https://www.numberanalytics.com/blog/control-theory-philosophical-lens-language
NCCR automation fellowship: https://nccr-automation.ch/research-fellowship
Andreas Krause: https://las.inf.ethz.ch/krausea
Steve Brunton ResNet video: https://www.youtube.com/watch?v=w1UsKanMatM&t=911s (https://www.youtube.com/watch?v=w1UsKanMatM&t=911s)
Original neural ODE paper: https://proceedings.neurips.cc/paper_files/paper/2018/file/69386f6bb1dfed68692a24c8686939b9-Paper.pdf
Speeding up Neural ODE training and inference: https://proceedings.mlr.press/v235/miao24a.html
Neural ODE review: https://www.sciencedirect.com/science/article/pii/S2950550X25000263
Review of Neural ODEs in biomedical engineering: https://www.annualreviews.org/content/journals/10.1146/annurev-bioeng-110824-124907
Neural ODEs for observer design paper: https://arxiv.org/abs/2212.00866
Non-linear observer review: https://www.sciencedirect.com/science/article/pii/S1367578821000948
Neural ODEs for control design papers: https://arxiv.org/abs/2504.17139
Survey on CBFs: https://coogan.ece.gatech.edu/papers/pdf/amesecc19.pdf
Koopman operator paper: https://arxiv.org/pdf/2509.07294
Combining neural ODEs with RL: https://www.sciencedirect.com/science/article/pii/S0925231225007131
Neural…