Type of placement
This was a research visit to work in collaboration with Dr Ricardo Martinez-Garcia at the Centre for Advanced Systems in Görlitz, Germany.
Andrei’s reflections on the placement
Much has been achieved with this visit and more is expected to come from it.
For the moment, I have achieved exciting results on multiscale pattern formation on rings that we intend to extend to infinite domains. This includes the development of new methods for stochastic and deterministic analysis of populations with diversity of interacting radii, which opens doors for analysis of other systems with eco-evolutionary dynamics and multi-dimensional trait spaces. Ricardo and I are planning to finish working on this project and write a publication to be submitted in the next couple of months. I have also submitted an abstract to a conference to talk about this project.
In the process of obtaining results for my project with Ricardo, I have developed many other skills. This includes new analytical methods for the analysis of stochastic and deterministic systems (analysis of pattern formation) as well as other skills in numerical methods such as coding in C++/Fortran and the use of pseudo-spectral methods to solve partial differential equations numerically. I am also using their cluster to run code (which having used BALENA in Bath before has helped me with).
Additionally, I was asked to give a departmental seminar at CASUS and met academics afterwards to discuss my research and mathematical tools they had that will be useful for future research I might do. On the networking side, I have exchange knowledge on analytical tools and coding libraries for data analysis with other PhDs/postdocs in Ricardo’s group and the department.
My current plans are to keep working with Ricardo on this project, write an article for publication, and potentially work on collaborations with him and some of his collaborators afterwards.
Overall, the placement was highly beneficial for both my PhD and my career as a whole. It not only opened doors for me and my supervisors to collaborate with Ricardo and other academics at CASUS, but it also gave visibility to SAMBa and the University of Bath.
I’d highly recommend 2-3 months research visits to both students and supervisors. The duration is sufficient to kick-start a project that has the potential to turn into a paper, a collaboration, or grant proposal, whilst not too long to be detrimental to the student’s PhD research. During the placement, the student will be exposed to several networking opportunities and probably be invited to give a talk. All of these are highly valuable for potential job applications afterwards. It also gives the opportunity to get to know a place or research group. For supervisors, the student will be exposing their research to different institutions and groups, providing good visibility that might result in citations, collaborations, and grant proposals.
All in all, it has been a fruitful experience, and I’d be happy to recommend it to both students and supervisors.
Comments from Andrei’s supervisor:
This placement has been highly beneficial to Andrei in helping him to build his research independence and international networks.
The whole process worked very smoothly from my point of view, but that was mainly because Andrei organised everything for himself.
I would certainly recommend any student to explore opportunities like this in the future.
Type of internship
3 month secondment with Mediatek on Machine Learning, particularly on ways to deal with scarce amount of data. I had the opportunity to study an algorithm called MAML and to extend it.
Paolo’s reflections on the experience:
I achieved a preprint and many research ideas. My research will benefit from this because I experienced a different and differently organised environment, with other ways (but similar) of working together to achieve the desired goals. In the near future, the document which I contributed to will be finished, so there will most probably be other interactions with people I worked with.
I recommend this kind of experiences to anyone interested in the subject.
Comments from Paolo’s supervisor:
The placement gave Paolo the possibility to get hands-on experience in working with neural networks, both by developing theory and by actually writing code. I believe that this ideally complements the deep foundational knowledge of Stochastic Analysis Paolo has already acquired in working on his PhD topic. I am confident that the combination of these skills will make him very attractive to future employers.
Paolo reported that he had daily meetings with one or two supervisors during this internship giving him immediate feedback on any progress made and thus accelerating the learning process.
Overall, I think that this was a good experience for Paolo and I would definitely recommend this to interested future students.
What did the internship involve?
Reporting to Nitin Patel (Co-Founder and CTO) and Jaydeep Bhattacharyya (Lead Software Engineer), I was required to understand and identify problems with a proprietary portfolio optimal decision-making software product, and to take the initiative in proposing and implementing solutions to these problems. This role was within the modelling and programming team.
Robbie’s reflections on the experience:
My communication and collaboration skills benefitted from this experience, and I learnt about the working environment within a private company in the software industry.
This role required having effective communication and collaboration skills to discuss potential approaches to the problem, and a strategic ability and knowledge of statistical software to implement appropriate solutions. Once solutions had been obtained, written and oral presentation skills were required to communicate them to appropriate stakeholders.
In this role, I successfully implemented an approach to extend an existing piece of software to allow for group sequential trials. In addition, with collaboration from Professor Christopher Jennison and other colleagues at Cytel, I developed theory and implemented an approach to compute ‘optimal’ group sequential trials.
We have plans to follow up the research in my doctoral studies and have future collaborations.
I would highly recommend interactions of this nature.
Comments from Robbie’s supervisor:
Robbie had a highly productive placement at Cytel and gained valuable experience working in a commercial organisation with a strong research agenda.
When I visited Robbie during his placement, we discussed with his manager an alternative approach (Dynamic Programming) to solve the optimisation problem posed. Since returning to Bath, Robbie has implemented this approach and shown that it speeds up calculations dramatically. This work has led him to extend his PhD research project in a new direction, implementing and applying this methodology.
I believe Robbie’s interaction with Cytel worked extremely well. I see the work-experience as valuable in itself and the spin-offs for Robbie’s PhD work were an added bonus. I would certainly recommend activities of this nature for our PhD students.
I worked at Roche as a PhD summer intern. I was part of a working group put together to develop guidance for statisticians on how to analyse trials in which treatment switching takes place. I produced a step by step guide that can be used by a statistician to analyse clinical trials in which treatment switching has occurred. This took the form of an R Markdown document and has been rolled out at Roche as part of a tool-kit. I worked primarily independently, supported by a small team, most of whom were based in Basel.
Lizzi’s reflections on the internship
This project greatly developed my knowledge of both basic and advanced survival analysis techniques as well as my R skills, particularly R Markdown. It was also a great networking opportunity and a chance to experience what it would be like to work for Roche as a company.
Since my PhD project is in collaboration with Roche, it was also a good opportunity for me to discuss my project with several of the Roche statisticians which gave some useful insight. I also presented on my PhD in a seminar to a wider group of statisticians when I visited Basel. Although the treatment switching project did not contribute to my research I had some useful feedback from presentations and some conversations on my PhD topic which have generated ideas. The questions I was asked at the end of one of my presentations also highlighted to me that I need to frame my research slightly differently to better convey the main message.
I have no formal plans for future engagement as a result of the placement but I already have a collaboration in place with Roche for my PhD project. It was also made very clear to me that they would like me to keep them in mind for my future career. It was a very positive experience and I would thoroughly recommend it to anybody interested in working as a statistician in the pharmaceutical industry.
Comments from Lizzi’s supervisor:
Lizzi had an excellent placement at Roche. Working in the Statistics Department of a major pharmaceutical company, she learnt a lot about the context of her PhD project. The project that Lizzi was engaged in involved topics of considerable current interest. There will be a more direct benefit to Lizzi’s PhD research from the new programming skills she has learnt. Also, Lizzi took the opportunity to discuss her PhD research topic with members of the Department and gained some useful ideas that will provide input into her work.
I believe the interaction with Roche worked very well and Lizzi’s internship strengthened the relationship with the Department at Bath. I would certainly recommend activities of this nature for our PhD students.
Type of internship:
12 month PhD Enrichment programme with The Alan Turing Institute
What did the programme involve?
The placement was called Enrichment student PhD programme, a program for PhD students from several universities in UK that would relocate for one year at The Alan Turing Institute in London. The students would continue their own research, however they would have the possibility to interact with many academic fellows from different universities and with potential industrial partners. On top of this, students had the possibility to try to join some research project of interest initiated by the Institute. Several conferences, seminars and data study groups were furthermore proposed, with the possibility for the students to broaden their knowledge in data science related topics.
During this year, I mostly continued doing my own research. On top of that, I was able to interact with many fellows, which gave me the possibility to embark on a Turing-led project about the construction of a 3D printed bridge currently under development in Amsterdam. My supervisor and another collaborator of mine were also able to join the project.
Gianluca’s reflections on the experience:
During this year, I met a lot of different people, academics, PhD student. I could network a lot for future possibilities and I made several new friends. It has been a fun and constructive year. In addition, being away from a close supervision for one year helped strengthening my discipline and independence. I could broaden my horizons and I got closer to the machine learning community, learning that their research can be very similar to mine in terms of methodology, but the purpose and the point of view may change slightly. During this year, two of my research papers were submitted and accepted.
I would definitely recommend such an experience. It has been a fun year, rich of new possibilities. The Alan Turing Institute is also a good institution to have on a CV.
Comments from Gianluca’s supervisor:
Gianluca has greatly benefitted from the enhancement year at the Turing Institute in London. The interaction during his time there has worked very well; instead of possibly meeting every week, we may have met slightly less frequently, but we spent more time together (than in usual weekly meetings).
The benefits from working in such an engaging, interdisciplinary, data-science focussed environment at the Turing Institute clearly outweighed any possible disadvantages due to us not being at the same location. In addition to Gianluca exchanging himself with other researchers at the Turing Institute and being able to go to seminars, workshops and other events there, we also got involved in a very interesting new project jointly with engineers and statisticians from other Turing partners, within the STEAM project, part of the data-centric engineering theme at the Turing Institute. This project is going to lead to future collaborations and is currently being written up for a paper in Nature Letters.
I was hired as a “Summer Placement – Engineering” student by CTT for 11 weeks. I worked on three different projects across two teams:
Kate’s reflections on the placement:
I gained insight into what it would be like to have a career in industry. It was great to experience the culture in the offices, and the learn about processes they use to progress research and development projects.
During my time at Cummins, we created a collaboration and they continue to help me with my research. I gained data to validate the current version of my compressor model. With the expertise of those in the company, I was able to identify some missing physics (diffuser stall) and started implementing this during my time there. This was completed once I returned to Bath, and a continued collaboration has further helped me in my research. Since the placement, I have been in email contact, and have been to visit them for a fortnight. They are currently supplying equipment to enable me to do some experimental testing for my PhD work. I hope to visit them again before the PhD is over.
I was also offered a sponsorship, where Cummins pay me some money during the final years of my PhD on the condition that I take a job with them once I finish. I have really enjoyed my interaction with CTT. They have been enthusiastic and supportive, and have helped me way beyond my expectations. I would definitely recommend future activities of this nature.
Feedback from Kate’s supervisor:
The secondment at Cummins was an extremely positive experience for Kate as it validated her research both giving her confidence that it was a practical model for applications, and providing data to calibrate her model. It opened up new questions and directions of research and provided her with additional support for the remainder of her thesis.
3 month placement with HSBC, Global Banking and Markets UK
What did the placement involve?
The three-month placement was at HSBC Global Banking and Markets, with FX eRisk Spot, quant strategies team in high frequency algorithmic trading of cash foreign exchange spot products. My project was to assess the performance of the market fill simulator for algorithmic execution of trades, and to propose, implement and test improvements. The project had a distinct interdisciplinary flavour, combining probability, statistics, complexity science and programming with applications to finance.
Anna’s reflections on the experience:
You can read a full report of Anna’s placement in blog post that she wrote for the University Careers website.
In a way, the format of the project was not too different from doing a PhD: it was technical, without a clear solution but with multiple possible ways to approach it. Beyond the technical knowledge and coding skills, the project relied on the ability to conduct research independently. However, the project was a lot more hands-on and goal-oriented with a clear deliverable objective at the end of my three-month internship.
Throughout the internship there were plenty of opportunities to learn, to get exposure and to present my work to senior management. Being part of a larger cohort of interns made the experience more sociable. Everyone was placed at different desks throughout Markets & Securities Services (MSS), and so we worked on a variety of real-life problems. This way, we could draw upon each other’s experiences and learn about the range of problems faced by quant teams and possible solutions suitable for some asset classes but not for others. We were also encouraged to shadow experienced colleagues and to attend networking sessions held over lunch, to learn more about various products and business lines within HSBC. Moreover, there were a few programming workshops organised for us, to help us pick up relevant coding skills.
Doing an internship during my PhD was a great way to learn about opportunities to apply my technical skills to high impact real-life problems in the finance industry and to secure a permanent role in the team I knew I enjoyed working with.
Comments from Anna’s supervisor:
The placement has helped Anna bridging the gap between research at university and research in industry. The team at HBSC directed her towards topics where she could use and enhance her expertise.
The experience was positive from my point of view and I would recommend similar activities at the final stages of a PhD if time permits.
Nature of the internship:
3 month secondment with the Department of Psychology at the University of Bath
What did the secondment involve?
The three-month secondment in the University of Bath was with Professor Julie Barnett and Dr Iulia Cioroianu analysing spatial data on mutual aid groups established in the UK during the COVID-19 pandemic and modelling the distribution of these groups in order to understand their effect on supporting connections between isolated individuals.
During the Coronavirus pandemic online groups, Mutual Aid Groups, formed to connect and support isolated individuals in the community. The aim of our project was to understand the relationship between the characteristics of the local authorities and the creation of the groups as well as the engagement of the community with these groups.
During the project I identified sources for and collated local-authority level data for a variety of variables from multiple sources in the UK and linked them with data for the Mutual Aid Groups into one comprehensive data set. With this data we fit spatial models to analyse the relationship between the creation and membership level of these groups with the characteristics of the local authorities and were able to analyse whether or not there was evidence that these variables had a significant effect on the engagement of the communities with the Mutual Aid Groups. The results from this project will be taken forward to a conference as well as submitted to a journal and there are many avenues for further research projects and collaborations.
Nadeen’s reflections on the experience:
Through this research project I gained further experience in data management which was applicable to my PhD work. The opportunity to work in an interdisciplinary team where I was communicating my work to people from different departments was invaluable for my work beyond my PhD, especially the chance to gain an insight into research in academia beyond my PhD.
I highly recommend taking on such a secondment, as the experience of research across multiple departments is incredibly valuable.
Comments from Nadeen’s supervisor:
During this secondment, Nadeen gained experience collating and modelling complex spatial datasets. She encountered and devised her own solutions to problem she does not face in her PhD research including changing local government boundaries over time and model selection. Nadeen has returned to her PhD research for several months now, and I can see that the experience tackling these problems in collaboration with a wider research team has given her new tools to reflect on and improve decisions she made in her research prior to starting the secondment. Nadeen shared her experiences with other PhD students through a presentation to my spatial statistics research group, and I believe it will shape her own career goals in the future.
Secondment with the UK Health Security Agency
I joined the UKHSA (formerly JBC) from March 22nd 2021 to March 31st 2022, during that time I worked within the Regional team, which was then merged into Regions and Sectors before ending up as the Thematic team.
Whilst there I worked on many topics, both close to the policy side and very far from the policy side. A few key highlights would be the creation of a tool utilising clustering algorithms which investigated regional heterogeneity within the country to identify outlying areas, a simple probabilistic model designed to identify risky settings, taking ownership of a specific metric and extending it to produce new outputs sent as part of a regular package to the Cabinet Office Task force. As well as many others. During my secondment I led on the production of 3 internal government reports highlighting results of my analysis and supported many other reports.
Jason’s reflections on the internship:
This secondment has given me the opportunity to widen my non-academic network as well as establishing new academic contacts. Through the placement I was invited to a weekly meeting with the Turing institute where I was afforded the opportunity to present my work and receive feedback. Beyond this I was also able to present to broader government colleagues on multiple occasions, sometimes presenting to an audience of over 100 people. Through the placement I have significantly developed the skills needed to be a data scientist, both supplementary tools such as git, and core skills such as data analysis, visualisation, and coding best practises. These skills will be valuable both during my PhD and beyond. During my secondment I received 2 peer-nominated “In Year Awards” for my contributions to the team and organisation, both as an analyst and as a colleague. Beyond this there are plans to convert some of the work I was finishing up towards the end of the placement into a short journal article, though the details are yet to be finalised.
Overall, I really enjoyed my secondment and would highly recommend them to anyone considering one. The opportunity to work in a fast paced, collaborative environment taught me a lot about both the internal workings of government and about best practises working in an agile environment. As far as advice goes for students considering a placement / secondment I would strongly encourage discussing ITTs they have attended, beyond this ensuring to touch base with their supervisors (and occasionally retouch some of their PhD work on a quiet day) to ensure the transition back afterwards is smooth.
