
Yoshua Bengio: equipping AI with higher level cognition and creativity
The Robot Brains Podcast
00:00
McMc - What's Next?
There are several ways to get maximum likelihood learning of an energy function. You can replace the MCMC by a G-Fluonet essentially. But there's an even more exciting approach where you don't even need to sample data at all. And another way that we've been working with is a Bayesian approach which I think is more like what the brain does.
Transcript
Play full episode