#66 – Michael Cohen on Input Tampering in Advanced RL Agents
Hear This Idea
00:00
The Differences Between Supervised and Unsupervised Learning
The way I've heard the difference framed or at least one version is like when you're doing supervised learning, you have a labeled dataset. When you're doing unsupervised learning, you don't. The label in that example would be like, is it a dog or not? So yeah, you could say that with unsupervisedlearning, there's just a set of data points, and so there's no labels. And when you're learning a function, you get the input value and the output value. That's why the difference is between learning functions and just like things. Yeah. There's a bit of a, when you describe it like this, you might notice weight functions
Play episode from 06:45
Transcript
