AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
3min chapter
Data Debt in Machine Learning with D. Sculley - #574
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Machine Learning Systems From a Datea Perspective
In general, really focusing on the discipline of model valuation is something that i think is incredibly important. There are some good ways to use the data that we already have to help probe and there are two that i quite like. We can first cluster our data before doing any training, and then do the moral equivalent of cross validation. But instead of holding hold out, we hold one of these clusters out. And so we're creating a world in which our models explicitly have some blind spots,. That blind spot corresponds to a cluster tha we then use as testato. If it's very far away from its nearest neighbor in the training data, then that's probably a pretty tricky example to get right
00:00
Transcript
Episode notes
Get the Snipd
podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any
moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share
& Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode