Provenance, security & machine learning
Monday 11 Nov 2019
Time: 09:45 - 17:30
Bio
Nick Holliman is Director of CUSP London and Professor in Computer Science at King's College London. He obtained a PhD from the Departments of Computing and Mechanical Engineering at the University of Leeds, sponsored by IBM UK. He worked in the computing industry, researching high quality graphics, computer vision and high-performance computing at Sharp's European Research Laboratories in Oxford, leading to a number of globally impactful products. He then moved to academia, most recently leading the development of research and teaching in the Data Science group at Newcastle University. His current research investigates the use of information theory to create novel visualization methods and empirically validates them worldwide using rigorous methods drawn from psychophysics.
He is one of the main organisers of the Visualization Turing Interest Group.
Research interests
His research aims to find new ways to create data visualisations that directly address challenges in the human understanding of big data and AI. Starting off in electronic form in a computer, a display device converts information optically to light, this electromagnetic information is detected by the eye and the information finally exists as electro-chemical nerve impulses in the brain. How should we optimise this process so that we can quickly and accurately understand new ideas, and with these ideas take decisions about the world?
His research at Newcastle has focused on exploiting the power of cloud super-computing to create more effective visualisations of urban analytics data. Cloud super-computing allows three orders of magnitude more computing power to be applied to data visualisation problems than would be possible on even the most powerful desktop computer. This is allowing the exploration of visual approaches to representing data that have not been see before, one example is the TeraScope project that is producing a visualisation from city scale to room scale in a single image of one trillion pixels.
As a Turing Fellow he focuses on two directions of research: First to address the challenge of visualising uncertainty, how might we represent values and our confidence in values in immediately understandable ways. Second to address the challenge of how to automate visualisation, big data is too big for a human designer to confidently represent visually, how can we automate this process using stochastic optimisation techniques and perhaps also machine learning.
Achievements and awards
As well as patents and academic papers his research outputs have included a number of award winning short stereoscopic 3D films on the role of dark matter in the creation of the universe in collaboration with Professor Carlos Frenk and the Institute for Computational Cosmology at Durham University. He is a member of the Association for Computing Machinery, The IEEE Computer Society, the Royal Statistical Society and the Society for Imaging Science and Technology.
