Another MCMC Method. Gibbs sampling is great for multivariate distributions where conditional densities are *easy* to sample from.
To emphasize a point in the video:
- First sample is (x0,y0)
- Next Sample is (x1,y1)
- Next Sample is (x2,y2)
...
That is, we update *all* variables once to get a new sample.
Intro MCMC Video : https://www.youtube.com/watch?v=yApmR-c_hKU