inControl cover image

ep3 - Ben Recht: A tour of optimization, machine learning, and control

inControl

00:00

Is There a Robustness in Machine Learning?

The fix for this in practice is constant retraining and there's a lot of that. You do have to refine the models, you do have to retrain the models. They're very unstable to the ravages of time. Even with poor prediction, we can have stable and robust behavior by proper design and feedback design. That's one of the more fascinating under appreciated aspects of control.

Play episode from 25:43
Transcript

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app