Tue 9 Jan 2018 14:00 - 15:00 at Bradbury - SESSION III (invited tutorial + 1 talk) Chair(s): Cameron Freer

The TensorFlow Distribution and Edward libraries implement a vision of probability theory adapted to the modern deep-learning paradigm of end-to-end differentiable computation. We first introduce TensorFlow Distributions, an efficient low-level system for building and manipulating distributions. We focus on the non-obvious design choices in the library, paying particular attention to the Bijector abstraction, which supports composable volume-tracking transformations with automatic caching. We then provide an overview of Edward, a probabilistic programming system built on computational graphs and using Distributions as an efficient backend. In particular, we show how Edward and TensorFlow Distributions can be applied for expanding the frontier of deep generative models and variational inference.

Tue 9 Jan

Displayed time zone: Tijuana, Baja California change

14:00 - 15:30
SESSION III (invited tutorial + 1 talk) PPS at Bradbury
Chair(s): Cameron Freer Remine and Borelian
14:00
60m
Talk
Deep Probabilistic Programming: TensorFlow Distributions and Edward
PPS
15:00
30m
Talk
More support for symbolic disintegration
PPS
Praveen Narayanan Indiana University, USA, Chung-chieh Shan Indiana University, USA