[MINI] One Shot Learning
Data Skeptic
00:00
The Basic Principle of One Shot Learning at Work
One shot learning is a form of reinforcement learning. It's where the system has to not so much learn just what is in a given class or not, but also it has to learn under what conditions should ask for help. For example if you were trying to classify a rare disease, like a really rare disease, something like one in ten million people get. One shot learning is trying to get machine learning to do the same process. And treated a little bit special and learn from a very small number of examples. So they need to study one shot learning more.
Transcript
Play full episode
