Type of internship:
UKRI Policy Internship with the Government Department for Environment, Food and Rural Affairs (Defra)
What did the internship involve?
The UKRI Policy Internship scheme is a three-month internship to gain experience on policy making informed by research. I was working on a Defra (Department for Environment, Food and Rural Affairs) project. It involved a literature review, data analysis and modelling.
During the placement I worked on modelling a pest distribution to predict its suitability in areas where it is settled and expanding, as well as assessing the risk in the UK. It involved a lot of data acquisition, handling and analysis, modelling techniques and programming skills.
Some of the simulation results can be published as part of a Pest Risk Analysis (PRA). This consists of a process to decide if a pest or pathogen should be regulated, based on scientific or economic evidence. It also identifies ways of control measures against the organism. In addition to this, there is a plan to publish the results as part of a scientific paper.
Laura’s reflections on the internship:
This placement had a lot of benefits for me in both my research and personally. First, I improved my programming skills, learnt new methods and sources of data that could be beneficial in my current research. Second, through different opportunities I had, such as shadowing, I was able to explore different areas within the civil service, for example how priority policies are considered. Finally, I was able to meet amazing people from different backgrounds that helped me so much understanding the work and made me feel welcomed. I feel it was a great experience and I would highly recommend it.
Comments from Laura’s supervisor:
The placement was a complete success. It provided a “break” from research at the right time allowing Laura to return with renewed focus and perspective on her thesis. It helped that the work, while separate from it, used some statistical and programming tools useful to her current research. Laura is as a result more confident about her skills. I certainly recommend it to SAMBa students!
Nature of the internship:
Research placement at the University of Oxford with Professor Christina Goldschmidt and Dr James Martin
The current outcome of this work is a proof of a conjecture stated here. At the present time, Christina, James, and I are working on a paper for this proof which we hope to submit shortly. Furthermore, many ideas were generated on problems involving random matchings which I hope to pursue with Christina and James once the paper is submitted. On top of the research produced which will contribute to my PhD thesis, this placement has built a long-lasting collaboration for me which I believe will be extremely useful if I pursue academia in the future.
Outside of my research, I was able to attend the Oxford probability and junior probability seminars, even being given the opportunity to present a talk at the latter. These were fantastic opportunities to broaden my knowledge in my subject area, and learn to speak about my work in a friendly environment. I plan to attend both of these seminars in the future. Outside of maths, I joined a running group with other PhD students in the department which was (exhausting!) fun and something I had not done seriously before. I cannot be thankful enough to how welcoming not only my placement supervisors were, but the entire probability department at Oxford. I believe there is much crossover between the probability groups at Oxford and Bath, and would fully recommend future collaborations, such as placements.
Advice to PhD students considering a placement/secondment:
Comments from Chris’ supervisor:
I believe that this placement will strengthen Chris’ CV and make him more competitive for future postdoc positions, and even for permanent positions in academia later in his career. Indeed, the subject that Chris agreed on with Christina and James is related to his previous work on Pólya urns, but also significantly different, which will allow Chris to greatly expand his area of expertise in probability. This placement was also a great opportunity to work with collaborators that are not his supervisors and that belong to another institution than Bath, which again will be a strong asset when applying for further positions in academia.
Research visit to Graz University of Technology, Austria.
The research visit was spent pursuing a new research project proposed by Thomas Pock on the topic of bilevel learning – the primary focus of my PhD thesis. The project consisted of comparing different approximations of the bilevel learning problem and making connections both between them and to the original optimisation problem.
Seb’s reflections:
The SAMBa Internationalisation fund has made this international collaboration possible that we are optimistic will lead to a publication in the future. I also presented my recent work to Thomas Pock and his group – the vision, learning, and optimisation (VLO) group – which garnered insightful discussions regarding future directions of research. Indeed, both Pock and his group are world leading in the topic of bilevel learning. This research visit allowed me to establish a strong rapport with the VLO group by networking with the group every day for the weeks of the research visit – which was only possible due to the in-person nature of the visit. As a result, this brought both the VLO and my supervisors’ groups closer together which could foster future collaborations between the two.
As I am determined to remain in academia after my PhD, this research visit provided me with an invaluable opportunity to work and network with world leading experts in the domain of computer vision and bilevel learning which will help lead to future career prospects. Living outside the UK for several weeks improved my own independence, a crucial skill from both a personal and academic perspective. I really enjoyed the research visit and am very glad that I pursued this opportunity – I would certainly recommend that other students consider doing them. For more details, see Seb’s blog: My research visit in Austria.
Comment’s from Seb’s supervisor:
In my view, Seb’s placement was a big success. He not only learned a lot academically but this research visit also supported his independence and wider outlook on the research landscape in a way which would not have been possible while staying in Bath. Since the hosting group was carefully chosen, Seb’s research over the summer directly feeds into his thesis. Also, this research stay gave Seb a great chance to network with world-leading scientists in his field, thereby directly supporting the next steps in his academic career.
When reading this form as a whole, it is apparent that this placement was money and time well spend and I am totally supportive of similar future activities. In fact, Seb’s placement motivated me to apply for an EU networking grant such that other PhD students can experience similar placements in the future without relying on internal funding.
What did the secondment involve?
During the secondment I built a generative model and a shape-based classifier for capturing variations in shape of hips with Dysplasia, one of the main risk factors for Osteoarthritis. This will potentially lead to a publication and can be a useful tool for the doctors with whom we work. This model can be useful for understanding modes of shape variation in hips, due to Dysplasia, as well as for a screening/risk assessment tool for assessing scans of patient’s hips, who may have Dysplasia and are at risk of developing Osteoarthritis.
Allen’s reflections on the experience:
In terms of the benefits to me and my research, I developed further skills in Pytorch, and learned new frameworks such as PyKeops, to build algorithms for large scale shape processing, with extremely high dimensional data. As a benefit to my research, the frameworks and ideas I learned about during the placement were quite useful when translated to my work, and allowed me to scale up and improve my own algorithms. This also helped to formulate some interesting questions for my own research, through thinking about the problems encountered during the placement.
We plan to publish the outcome of the collaborative work done during the placement in an orthopaedics related journal, and I am open to collaborating on any further research directions that grows out of this project.
Working at the IMI was a lovely experience, and the team really made me feel welcome and helped settle into the new role. There was also a chance to learn about the other projects happening there, and chance for discussion and advice on your own research from the MIRAs. I would absolutely recommend such an experience, as it not only allows you to widen your research background, but also to get an understanding of how valuable mathematical research can be in such a diverse set of problem domains, many of which I would not have been aware of prior to this.
UKRI Policy Internship with the Government Office for Science
During my internship I conducted desk research in cross-cutting and fast paced areas of international policy, including compute and data sharing/access for Artificial Intelligence. This involved gathering evidence/information from different complex sources, such as government strategies, reviews, and articles, to produce literature reviews and gap analyses which summarise the policy landscape in the UK and the US. The information was vital for informing further briefings and information sheets, which I led on developing. My outputs ensured that council members were thoroughly briefed on these policy areas ahead of council wide meetings with the head of research at Google and the director of the EPCC, leading to an insightful discussion. My information sheet was also critical in highlighting areas of shared US/UK interest, guiding CST and PCAST members to have the most impactful discussion points at their joint meeting.
I assisted with drafting an advice framework from the CST on compute for AI. I used desk research and prior knowledge of the topic to clarify the questions to the secretariat team and identify areas of the commission that required further information to answer. My reflections on the secretariat’s stakeholder engagement and evidence gathering were important to guide the structure and content of CST’s advice, and gave helpful steers on how to make the recommendations clearer and more accessible. I sat in on fast paced and high-level meetings with various stakeholders, and produced formal records of the meetings, capturing complex and nuanced discussions. Throughout my internship I worked flexibly and across the team, adapting to and assisting with projects outside my immediate remit when reactive support was required. This required me to work at pace, often with short-term deadlines.
Rosa’s reflections on the internship:
Through the internship I gained insight in science-driven policy and decision making, and learned first-hand about the role and ways of working of the Council for Science and Technology. The experience manifested my aspiration of a career at the intersection of science and policy and is highly valuable for future job applications. I highly recommend the internship to PhD students who are curious about research impact and policy and who are looking to gain work experience beyond academia and industry. I believe this experience will be valuable both for students aiming to stay in academia and to those who are considering other career paths.
Comments from Rosa’s supervisor:
The Internship was an excellent opportunity for Rosa to engage with the process of converting research outputs into policy. Rosa gained substantial experience and transferable skills and training relevant to the future career paths that are likely to involve science as well as policy.
