Top Traders Unplugged cover image

TTU132: Trend Following Evolution ft. Ed Tricker, CIO of Quant Strategies at Graham Capital Management

Top Traders Unplugged

00:00

Discrete Problems Are Better Than Continuous Problems

A lot of these techniques fall under the broad world of reinforcement learning. They say, today, I'm going to set my valve here and this is the amount of days it's coming in. Comes in and the model speeds up. And then tomorrow, they'll repeat. That is the process by which these models will update their view. It is much better to do things in a continuous fashion than it is to doThings in a discrete fashion. Discrete problems mean that a very small change in what goes in can lead to a very big change in what comes out.

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