MLOps.community  cover image

Data Selection for Data-Centric AI: Data Quality Over Quantity // Cody Coleman // Coffee Sessions #59

MLOps.community

00:00

Insuring Data Quality at Every Step of the Way

Data is the holy grail for machine learning, but it's still a black art from an academic perspective. Data formas are more nacent as far as how people think about it and how they can manage it. You should be inspecting kind of the data kind af out each point to make sure that it's like, ok, what's going on? What's the quality? Where are the issues and stelf like that. Bad data at the very beginning just has this kindlike amplifying effect and can just pollute everything. Your test set is constantly changing in any given context. In any given moment you might change the rules right off the bat. For example, we work in

Transcript
Play full episode

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