Machine Learning

The traditional way of training models is quite time consuming and slow or at least I think of it as being quite time consuming. And it's surprising to me that that sort of approach would work in a financial situation because the market can switch and change relatively quickly. Different way is to have the model be trained on the subset of data with a time horizon and a window that actually evolves based on the time you're targeting, time horizon you're targeting. So that's called streaming. That's data stream algorithms and that's the one we presented with Albert of the MOA or Samoy from Apache.

Transcript

Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!

Get the app