80,000 Hours Podcast cover image

#48 - Brian Christian on better living through the wisdom of computer science

80,000 Hours Podcast

00:00

How Does Upper Confidence Bound Work?

The way the strategy would work is like an suppose you have a process which is like, 80 percent of the time, i'm going to pull ta thing that i think te pull. And 20 percent of thetime, which ebsalom, you're going to just pull around em liva,. And then you slet you decrease that percentage as you go on. So one of the other strategies that's very simple and intuitive, but also offers this perty of log rythmic regret, is what's called upper confidence bound. It says we are not actually inte sted in the expected value of the machine and we're not interested in the lower bound. We are only

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app