Timezone: »

 
Poster
Storchastic: A Framework for General Stochastic Automatic Differentiation
Emile van Krieken · Jakub Tomczak · Annette Ten Teije

Thu Dec 09 08:30 AM -- 10:00 AM (PST) @
in Poster Session 6 »

Modelers use automatic differentiation (AD) of computation graphs to implement complex Deep Learning models without defining gradient computations. Stochastic AD extends AD to stochastic computation graphs with sampling steps, which arise when modelers handle the intractable expectations common in Reinforcement Learning and Variational Inference. However, current methods for stochastic AD are limited: They are either only applicable to continuous random variables and differentiable functions, or can only use simple but high variance score-function estimators. To overcome these limitations, we introduce Storchastic, a new framework for AD of stochastic computation graphs. Storchastic allows the modeler to choose from a wide variety of gradient estimation methods at each sampling step, to optimally reduce the variance of the gradient estimates. Furthermore, Storchastic is provably unbiased for estimation of any-order gradients, and generalizes variance reduction techniques to higher-order gradient estimates. Finally, we implement Storchastic as a PyTorch library at github.com/HEmile/storchastic.

Keywords: Deep Learning · Optimization · Reinforcement Learning and Planning · Generative Model · Graph Learning

Author Information

Emile van Krieken (Vrije Universiteit Amsterdam)
Jakub Tomczak (Vrije Universiteit Amsterdam)
Annette Ten Teije (Vrije Universiteit Amsterdam)

More from the Same Authors