Foundations of Data Science

Co-Leads: Nial Friel and James Gleeson

We consider a challenging research agenda under the broad umbrella of ‘foundations of data science’. This programme of research requires the development of synergies between mathematics, statistics and computer science.

One such is at the interface of applied mathematics and statistics. For example, modelling processes on networks; eg retweeting behaviour or epidemic spread; have been well studied by applied mathematicians. These are usually deterministic in nature. Here we aim to leverage this work by developing statistical methodology to allow parameter and model uncertainty.

Several emerging and important issues around the practice of data science have emerged around the issue of trust and opacity of AI systems. These are arguably the main block to their widespread adoption in enterprise, health and other important societal areas and also require a multidisciplinary input in order to realise these challenges. For example, we will explore fundamental research to develop general solutions to explainable AI and also to question around the ethical use of AI.