The Digital Futures team are hosting talks by The University of Manchester's new Turing Fellows for 2021-22.
Presenting at this event are Yu-wang Chen and Matthew Thorpe.
To register for this event, please visit the Eventbrite link opposite.
Yu-wang Chen. Yu-wang is Professor in Decision Sciences and Business Analytics at Alliance Manchester Business School (AMBS). His research focuses primarily on decision sciences and data analytics, including their applications to risk analysis, supply chain management, consumer preference prediction, healthcare decision support, etc.
Yu-wang's talk is titled 'Decision analytics using data-driven modelling and evidential reasoning'
Decision analytics allow individuals and organizations to transform data and combine evidence to support informed decision making. However, real-world decision making problems are often characterized by multiple sources of data and different types of information. In this talk, I will briefly introduce my research on data-driven modelling and evidential reasoning in the context of decision analytics under uncertainties, and illustrative examples will be used from the fields of both engineering and management.
Matthew Thorpe. Matt is a senior lecturer in Applied Mathematics. His research focuses on the application of methods from the calculus of variations to analyse large data limits of problems that arise from machine learning and statistics, and in the application and development of optimal transport distances to signal and image analysis.
Matt's talk is titled 'Large Data Limits in Semi-Supervised Learning'.
I will give an introduction to my work focussing on Laplace learning, a semi-supervised graph-based methodology. In particular, I will state the Laplace learning method and show some large data limits which will identify the asymptotically well-posed and ill-posed scaling regimes. By making connection with random walks I will show one example of where bias in the asymptotic ill-posed regime can be corrected which leads to a new algorithm called Poisson learning which has state-of-the-art performance for low label rates. I'll also state some other projects that I am involved in which include the analysis of medical images and the linearisation of optimal transport distances.
