Gradient Dissent: Conversations on AI cover image

Shreya Shankar — Operationalizing Machine Learning

Gradient Dissent: Conversations on AI

CHAPTER

Is Distribution Shift a Problem?

I used to believe that this baby is a problem depending on how you define it. But the idea that if you have this static model in production, like a static set of parameters or a function that's being called on some features and these features are changing. At some point, right, you're that this is like the classic view staleness problem in database,. You need to refresh your views to keep up with the underlying data. And one concrete example of this is distribution shift.

00:00
Transcript
Play full episode

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
App store bannerPlay store banner