AI Today Podcast: AI Glossary Series – Dimension, Curse of Dimensionality, Dimensionality Reduction
AI Today Podcast
00:00
Machine Learning Is Data Hungry
Each dimension is going to add greater value in terms of differentiation, pattern discovery as well. But it's also going to make things more complex because we're adding more data. Typically, a rule of thumb is that you're going to want to have at least five training examples for each dimension in the representation. This is why machine learning is data hungry. The more you start adding, the more you just start introducing outliers and so on.
Play episode from 06:02
Transcript
