
Episode 493: Ram Sriharsha on Vectors in Machine Learning
Software Engineering Radio - the podcast for professional software developers
00:00
The Biggest Challenge With Distance Functions
Exact nearest neighbor search is just computationally infeasible. Approximation itself has challenges in high dimensions. Some distant metrics may not have good approximation guarantees for a generic metric. But usually the problems are even even earlier than that.
Transcript
Play full episode