Speaker: Jamie Townsend; Research Student, University College London
Title: Automatic differentiation
Abstract: In this talk I’ll present the basic methodology of automatic differentiation (autodiff), give an overview of existing software packages and outline some creative uses of autodiff, from tuning learning rates of neural networks to automatically generating code for inference in exponential family graphical models…
Speaker: Adam Lopez; Reader in Statistical Machine Translation, ILCC
Title: On the frontier of Natural Language Processing
Abstract: Natural language processing (NLP) has been enormously successful over the last decade, highlighted by offerings like Siri, Amazon Alexa, Google translate, and IBM Watson. What is left to do? I’ll discuss some problems that we don’t yet know how to solve. The talk will be very informal, covering things that we really don’t know how to do, and the ups and downs of getting started on a completely new problem.