Azure & DevOps Podcast cover image

Andy Roberts: Data in a Development World - Episode 243

Azure & DevOps Podcast

00:00

How to Predict Machine Learning Failures

With machine learning, of course, everybody's talking about GPT. It always strikes me that the P is pre-trained. Is MLOps partially taking a stream of events of new data and streaming it into the model for new training every day as new events happen? You might say that I'm looking at the probability that a machine is going to break down in some way. As you get better and better at that, you're actually starting to see the number of failures reduce.

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