Robert Eyre
I am currently a Data Scientist at Spectra Analytics. Spectra is a data science consultancy firm founded in the Complexity Centre, supplying bespoke data solutions to clients from a wide range of industries. I am part of the data science team responsible for analysing data and building models to address the diverse data needs of our clientele.
I completed a PhD in Complexity Science at the Complexity Science Doctoral Training Centre from 2014 to 2018.
My PhD research was in applied complex statistical models in public health, supervised by Frances Griffiths of the Warwick Medical School; Thomas House of the School of Mathematics, University of Manchester; and F. Xavier Gómez-Olivé of the MRC/Wits Agincourt Rural Public Health and Health Transitions Research Unit.
It is common in public health and epidemiology research to rely on a small library of highly standardised and established statistical methods. My work sought to bring more recent innovative methods that have seen common use in areas such as machine learning and computer science to public health research areas and datasets where these methods have seen little to no use. This involved statistical inference and computation on datasets such as the one compiled by the MRC/Wits Agincourt Research Unit, which has a Health and Socio-demographic Surveillance System (HDSS) and performs annual surveys in a rural but densely populated region of South Africa. Research projects included working on developing social contagion models, using Gaussian process regression in the calculation of fertility rates, and analysing household food security using Bayesian belief networks.
I set up and ran the Complexity Centre's machine learning reading group from 2014-2016. Go check out what they're up to now.
Research Interests
- Data science and analysis.
- Applied and computational statistics.
- Bayesian inference.
- Machine learning.
- Public health.
- Social science.
Publications
- Robert W. Eyre, Thomas House, F. Xavier Gómez-Olivé, and Frances E. Griffiths (2018) Modelling fertility in rural South Africa with combined nonlinear parametric and semi-parametric methods. Emerg. Themes Epidemiol. 15: 5. https://doi.org/10.1186/s12982-018-0073-y
- Robert W. Eyre, Thomas House, Edward M. Hill, and Frances E. Griffiths (2017) Spreading of components of mood in adolescent social networks. R. Soc. open sci. 4: 170336. http://dx.doi.org/10.1098/rsos.170336
- Ian J. Ford and Robert W. Eyre (2015) Work relations for a system governed by Tsallis statistics, Phys. Rev. E 92, 022143. https://doi.org/10.1103/PhysRevE.92.022143
Background
- Born and raised in Winchester, UK.
- 2009-2013 - completed a Theoretical Physics MSci at University College London with first class honours. Project: Entropy production in small systems.
- 2013-2014 - completed a Complexity Science MSc at the University of Warwick with distinction. Project 1: Social networks and health. Project 2: Interactions between colloidal knots in liquid crystals.
- June-September 2016 - completed a data science internship with Spectra Analytics.
Schools and Conferences
- Academy for PhD Training in Statistics, Week 3, Warwick, July 2015.
- XXXV International Sunbelt Social Network Conference, Brighton, June 2015 - presented a talk on Social Contagion Over Adolescent Friendship Networks
. - Division of Health Sciences seminar series, Warwick, 4th June 2015 - presented a seminar with Ed Hill on Social Contagion Over Adolescent Friendship Networks.
- Academy for PhD Training in Statistics, Week 1, Cambridge, December 2014.
- First International Conference in Systems and Complexity Science for Healthcare, Washington, D.C., November 2014 - presented a poster on Social Networks and Health: Spreading of depressive symptoms over adolescent friendship networks.
- Student Conference on Complexity Science, Brighton, August 2014 - presented a poster on Social Networks and Health: Spreading of depressive symptoms over adolescent friendship networks.
