Neural networks in NLP

In the past few years, neural networks have proved themselves indispensable for tasks in natural language processing (NLP), allowing us to solve tasks that would prove difficult with previous methods. In this talk, I will review how neural networks are used in NLP , with a particular emphasis on recurrent neural networks (RNNs) in natural language generation tasks. Finally, I will present recent work on augmenting RNNs to take into account linguistic structure.

• Speaker: Kris Cao (Computer Science / NLP)
• Tuesday 21 November 2017, 14:00–15:00
• Venue: Centre for Mathematical Sciences, MR2.
• Series: Mathematics and Machine Learning; organiser: Damon Wischik.