K-Means in Practice
Data Skeptic
00:00
Is There Any Data to Supervised Learning?
The bank did not have any, like, any past instances of a fraud. We were looking for uncipervised learning techniques. It's also worth mentioning that we had a lot of categorical data,. You move away from means and you go into modes and medeoids. And in our case, running just the first test of like a k meteoid or mods was immensely computationally expensive.
Transcript
Play full episode
