Mathematicians selected to present work at the House of Commons
Dr Peter Johnson, Jeffrey Chu, Massimiliano Fasi and Mante Zemaityte have been short listed for the 2017 STEM for BRITAIN competition and will display their posters at the House of Commons on Monday 13th March in the Mathematical Sciences Session
The overall aim of STEM for Britain is to encourage, support and promote Britain's early-career scientists, engineers, technologists, and mathematicians who are the driving force of the UK’s research and development. These researchers are a vital asset for the future of the UK.
During this event in March 2017, there are 3 poster exhibitions and judging sessions during the day, each ending with a reception and prize-giving. The competition currently attracts around 500 entrants, of whom approximately only a third are selected to present their work in Parliament.
STEM for BRITAIN Awards are made on the basis of the very best research work and results by an early-stage or early-career researcher together with their ability to communicate their work to a lay audience.
Here is a summary of each researcher's work to be presented:
Dr Peter Johnson, an EPSRC Doctoral Prize Fellow, will be presenting his poster at the Mathematical Sciences session. His poster is entitled ‘Optimal Detection in Orbit’, outlining his research in Applied Probability and a recent application in Global Positioning Systems. Optimal detection of changes in random processes is important in a wide array of applications including finance, cyber security, epidemiology, meteorology, and statistical diagnosis. The optimal detection method used means that no other method will be able to detect the change as quickly and accurately as the algorithm being developed.
Jeffrey Chu is studying for a PhD in the research area of Statistics. His poster's content is about estimating the population size of big data sets. For example, in the analysis and modelling of hard-to-reach populations - i.e. terrorist groups, social networks, computer network traffic data, satellite and wireless communications/infrastructure etc. Often one of the first tasks in this type of analysis is to calculate the population size of such groups, whether it be the number of individuals, users, or some other quantity. He is helping to develop an estimator based on random walk sampling which performs well with small samples of big data.
PhD Student Massimiliano Fasi will be presenting a poster entitled "Finding communities in Large Signed Networks with the Weighted Geometric Mean of Laplacians". In network analysis, the detection of communities, that is, groups of nodes closely related with few interactions with the rest of the network, is an ubiquitous task. Obtaining a meaningful result is particularly challenging when dealing with signed networks, where beside positive relations such as agreement and similarity, one also considers negative ones. The poster discusses the advantages and drawbacks of performing this operation by using the weighted geometric mean of two matrices.
Mante Zemaityte will be presenting a poster as part of their PhD work entitled 'A Shift-and-Invert Lanczos Algorithm for the Dynamic Analysis of Structures'. The poster will showcase new ideas brought to some well established algorithms used by structural engineers for finding the natural frequencies of buildings and bridges. This work is crucial to avoid the phenomenon of resonance which may be induced by pedestrians walking over a bridge, or in earthquake prone areas.