The Bayesian Crowd
Speaker: Professor Stephen Roberts
Date: 4 July 2017
Time: 18:00 – 20:00
Watch the live stream from 18:30 – 19:30
The Alan Turing Institute is collaborating with G-Research to launch a new event series around the subject of data science, its theories and applications.
Researchers from the Institute will deliver six technical talks on their research and ideas – ranging from machine learning and artificial intelligence to new ways to make sense of data – which will be live-streamed online and afterwards made available to watch on YouTube and other platforms.
The first speaker will be Professor Stephen Roberts, Turing Faculty Fellow and Professor of Machine Learning at the University of Oxford, on ‘The Bayesian Crowd’, on 4 July 2017.
Professor Roberts will discuss how in applications such as crowdsourcing, a large amount of data needs to be combined in an intelligent manner and looks at how Bayesian models can help do this.
For realistic deployment, methods for this kind of combination should conform to optimality wherever possible, and yet scale well with large numbers of information sources and large amounts of data. This talk focuses on Bayesian information aggregation models. Professor Stephen Roberts will discuss how the use of approximate inference, based on variational learning, allows excellent scaling properties and retains high performance. He will showcase the breadth of applicability of the approach with examples from large crowdsourcing domains, fusing disparate information in finance, as well as feedback and user-task allocation mechanisms.
Subsequent speakers will be announced over the coming months and the talks are expected to take place between June 2017 and June 2018.
The talks will be delivered to G-Research employees and Turing researchers. Further speakers will be announced in the coming months.