Machine Learning for High-Risk Applications
The Data Exchange with Ben Lorica
00:00
The Importance of Data in Engineering Design
A lot of, so models have weak spots where they perform a logic and make silly mistakes. And it's typically because we're either out on the tails or just for whatever reason, we're operating in a sparse part of the training data. If you don't take anything else out of this podcast, check your model for weak spots in performance across segments.
Play episode from 28:50
Transcript
