Rating is available when the video has been rented.
This feature is not available right now. Please try again later.
Published on Nov 9, 2011
This video shows stochastic gradient descent with an annealing stepsize. This algorithm is efficient because it only uses a small subset of samples (perhaps even just 1 sample) to make updates. However, it will only find the state that maximizes the probability. It does not sample from the posterior distribution.