Approximate Inference and Deep Generative Models

About

Advances in deep generative models are at the forefront of deep learning research because of the promise they offer for allowing data-efficient learning, and for model-based reinforcement learning. In this talk I'll review a few standard methods for approximate inference and introduce modern approximations which allow for efficient large-scale training of a wide variety of generative models. Finally, I'll demonstrate several important application of these models to density estimation, missing data imputation, data compression and planning. .

0:00:00 Slide 1

0:01:59 Slide 2

0:04:21 Slide 3

0:05:49 Slide 4

0:07:06 Slide 5

0:07:29 Slide 6

0:07:39 Slide 7

0:07:46 Slide 8

0:08:29 Slide 9

0:08:36 Slide 10

0:09:50 Slide 11