[24] Martin Arjovsky - Out of Distribution Generalization in Machine Learning
The Thesis Review
00:00
The Thome Transversality Theorem in Differential Topology
It was kind of the coming down of being able to comfortably handle these things. The main theorem that i used was this tool in differential topology it's called the Thome Transversality Theory. It basically says that if you have manifolds that are random, they intersect transversally rather than parallelly like two parallel lines or six seconds.
Transcript
Play full episode
