Dr Rudkin’s research takes cutting edge techniques from econometrics and data science and applies these to problems in Economics and Finance.
Recent publications in Environment and Planning A and Food Policy document how taking a distributional regression perspective can identify crucial inequalities in the impact of supermarkets as food environment interventions. These works demonstrate lower price and wider product ranges simply reinforce behaviour amongst those with the poorest diets and that rather than simply focus on average diet improvements policy-makers should instead concentrate on how to aid those whose health needs the intervention. Theoretical work on this topic appeared in Economics Letters.
Understanding the benefit new techniques can bring to established wisdom is also highlighted in the power of the topological data analysis lens. In Expert Systems with Applications Simon’s work showcases how viewing a full data space directs attention to the true cases of interest. This is an applicability which occupies much of his ongoing agenda.
Explore Dr Rudkin's research and findings in Are supermarkets good for your health?, which is featured in Swansea University’s Exploring Global Problems podcast series.
Present research projects exploit the ability of Topological Data Analysis to present the true shape of data, giving the user new insight into what is really contained within the data they hold. Enabling the visualisation of complex multidimensional data sets with the ease of interpretation of a scatter plot, current papers demonstrate how combinations of data characteristics can yield outcomes of concern that seeking significance in regression approaches might miss. For example, Brexit support was concentrated in a small part of the voter characteristics space, whilst Remain spread far wider with associated challenges for creating unifying narratives. Likewise, firm failures occur in a far smaller part of the financial ratios space than the linear models used in credit risk appraisal would suggest. This work takes from the natural sciences and is pioneering within Economics and the social sciences but offers huge potential impact. Simon is very happy to discuss opportunities to unlock the information within any datasets.
Topological data analysis also has strong resonance in Finance where the dynamics of agent interactions become higher frequency and more noisy than they are in the traditional economy. Early work highlights how capturing disturbances to data may forewarn of impending crashes and how detecting periodic behaviours in the topology of time series can further enhance detection.