Student Stories: PhD student Jordan Robinson shares his experience of settling into the CDT during the pandemic
CDT sudents collaborate on various research-related mini-projects (aka hackathons for those with a more competitive streak!). This blog shares an insight into one student's experiences of settling into the CDT (during a pandemic) and his experience taking part in his first CDT mini-project.
My Journey
I graduated last Summer from the University of Liverpool with a Masters in Mechanical Engineering. I’ve always found machine learning fascinating and have conducted a variety of different projects over the last few years. For example, I created a line-following robot using a camera connected to a Raspberry Pi and analysed numerous datasets that I found on Kaggle.
In September 2020, I started working full-time as a Graduate Mechanical Engineer. Not long after starting, I realised that job wasn’t for me and decided to apply for a studentship at the CDT. Thankfully, my application was successful and, since starting in December, I’ve not looked back!
My Project: Distributed Hypothesis Generation and Evaluation
My research involves using machine learning to aid analysts in complex decision-making. Making the correct judgement requires an analyst to derive all of the potential hypotheses for a given situation. This is quite challenging and is the main focus of my current research. After this, it’s just a matter of proving or disproving the generated hypotheses (which I’m sure will be a lot harder than it sounds).
My research will benefit domains where the speed of analyses and validity of conclusions is imperative, and applies to a variety of fields including intelligence, adjudication in legal cases, and supporting company’s during strategic planning meetings.
My Experience at the CDT
Contrary to what you may think, joining the CDT during the pandemic has been a breeze! Although I’ve not actually met any of the research group in person yet, I feel very much part of the community thanks to friendly colleagues (and, of course, Microsoft Teams).
The expansion of my knowledge over the past 4 months has been remarkable, and I can directly attribute that to being part of a cohort within the CDT. The weekly CDT meetings are great for learning about other students’ research as well as hearing updates on potential group projects for the future. There is also a monthly journal club where we discuss the latest AI-related papers, as well as workshops with other machine learning research groups from around the world, most recently with Fraunhofer FKIE.
Students at the CDT are from a variety of academic backgrounds. The CDT provides cohort- and project-specific training to ensure students have all the necessary skills to complete their research. The advantage of having students with diverse academic backgrounds is that it enables us to think creatively as a team when solving challenging problems.
Solving real-world problems is at the very heart of the CDT and, as such, a variety of mini-projects are run throughout the year to help develop students’ skills and relationships.
Mini-Project: Clustering Types of Behaviour using Data from the University of Liverpool Libraries
Last week, all CDT students banded together for one day to conduct some exploratory data analysis on a number of different datasets given to us by the University of Liverpool’s libraries. The data being analysed was from the 2018 – 2019 academic year. The datasets contained information such as: which library they attended; the time of their arrival and departure; and whether they withdrew, returned or renewed a book loan. The datasets also included a demographic profile of the students, containing information such as: which department they were in; their year of study; where they were from; and whether their parents had attended university. From this, the library staff wanted to cluster types of behaviour exhibited by people using the library, with a view to tailoring the marketing of the libraries with said clusters.
The project was a blast and everyone from the CDT enjoyed the experience – so much so that some of the students have continued their analysis of the datasets even as I’m writing this post! The project has developed my communication, teamwork, and exploratory data analysis skills. However, the main takeaway for me is that it has furthered relationships between students, drawing people together and making the CDT a tighter-knit community. I thoroughly enjoyed the experience and I’m looking forward to taking part in another mini-project in the future.
Jordan Robinson, Cohort 2 PhD working alongside Dstl