Abstract
Derivatives, mostly in the form of gradients and Hessians, are ubiquitous in
machine learning. Automatic differentiation (AD), also called algorithmic
differentiation or simply äutodiff", is a family of techniques similar to but
more general than backpropagation for efficiently and accurately evaluating
derivatives of numeric functions expressed as computer programs. AD is a small
but established field with applications in areas including computational fluid
dynamics, atmospheric sciences, and engineering design optimization. Until very
recently, the fields of machine learning and AD have largely been unaware of
each other and, in some cases, have independently discovered each other's
results. Despite its relevance, general-purpose AD has been missing from the
machine learning toolbox, a situation slowly changing with its ongoing adoption
under the names "dynamic computational graphs" and "differentiable
programming". We survey the intersection of AD and machine learning, cover
applications where AD has direct relevance, and address the main implementation
techniques. By precisely defining the main differentiation techniques and their
interrelationships, we aim to bring clarity to the usage of the terms
äutodiff", äutomatic differentiation", and "symbolic differentiation" as
these are encountered more and more in machine learning settings.
Users
Please
log in to take part in the discussion (add own reviews or comments).