Wed 26th April 2017 – Dr Richard Dybowski
Dr Richard Dybowski – Disease Dynamics Unit, University of Cambridge
Two recent developments in neural computation: NTMs and GANs
The presentation will focus on two recent developments from the neural computing community. The first is the concept of the Neural Turing Machine (NTM), which is a fully differentiable system with the potential to learn algorithmic solutions to problems, and which has outperformed LSTM-based recurrent networks. The second is the Generative Adversarial Network (GAN), in which the parameters of a probabilistic generative model are estimated via a minimax game involving a discriminating deep neural network.
I am an applied mathematician, statistician and computer scientist based in the Disease Dynamics Unit, University of Cambridge. Following a degree in chemistry, I specialised in NMR spectroscopy. I undertook a PhD in statistical and computational chemistry at Leeds University (i.e., the computational diagnosis of endocrine metabolites) and was then appointed as a Research Fellow at St Thomas’ Hospital/King’s College London, where I undertook statistical research into a variety of medical areas, including microbiology, intensive care, and ophthalmology. I also taught artificial intelligence (AI) as a University Lecturer in computer science at the University of East London. After a number of years away from academia caring for others, I returned in earnest.
Venue and Time
Lecture Room N117, 14:00 – 15:00.
School of Computing Science & Digital Media, Robert Gordon University, Sir Ian Wood Building, Garthdee, Aberdeen, AB10 7GJ