Partner engagement – working with Improbable

One of the first things I learnt on this PhD programme, was understanding that I was in a partnership – not only with my supervisors and the university, but also with my Sponsor. In the first year of our programme, we all undertook an internship module with the objective to work alongside our sponsors on a small project. And so, for two weeks I had the opportunity to work with Improbable, out of their London office.

Before I continue, I should properly introduce Improbable to those who may not know who or what they are. Improbable is a British Multinational technology company that was founded in 2012. They created SpatialOS, a platform at which you can perform large-scale simulations and create virtual worlds and environments, for uses not limited to, video games and corporate simulations. They have a history of partnerships with Google, Softbank and Epic Games.

And so, learning all this about them and what they do, I knew I had to maximise my experience with them by gaining as much knowledge and skill from them as possible.

Whilst with them in London, I got to learn the projects in development, their areas of interest and how I fit into the grand scheme. Another became conscious of was, when having discussions about research, regardless of who with, transparency and respect are integral to the betterment of that relationship – and so in the context of my relationship with Improbable, I was glad to have a clear, open and professional channel of communication with them throughout the internship and afterwards.

A few months later, following a successful two weeks with Improbable, they sent Katie who official job title is Applied Scientist, to work with us out of LIDA for two weeks. At the start of her two weeks here, we introduced her the other PhD students and to some of the research going on within our CDT. Then, over the following days she worked closely on a project with my supervisors, which she was kind enough to summarise as the following:

I was investigating dynamic fidelity – which can be described as the ability to switch between an ABM and a higher-level, less computationally expensive model – in simulations of civilian movement around a city. The high-level model would be learnt from simulating using the ABM for an initial period before switching to the high-level model. The simulation would continue to run using the high-level model until there was some material change in the environment, causing civilian patterns of movement to significantly change in some way and resulting in a switch back to the ABM.”

There are quite a few PhD students, including myself, that are using ABMs i.e. Agent-based modelling methodology within our projects, so what was helpful about Katie’s work was the broadening of my understanding of ABMs as a tool.

On reflection of my first year on this programme, some of my most interesting experiences have occurred whilst collaborating with my sponsor. Though working in partnerships and in teams can be challenging at times, from brainstorming ideas and intense debates, the knowledge and the friendships gained are invaluable. Over the next academic year, Improbable are happy to give me another opportunity to intern with them again, but this time for much longer than two weeks. I am truly excited to see what we get up to next.

 

This article was written by Deborah Olukan, one of our second year students based in Leeds. You can view her full profile here.

Centre for Data Analytics and Society welcomes its third cohort of students

Hi! We are the newest University of Liverpool based cohort, part of the third year intake for the CDT in Data Analytics and Society. In our group we have James, Cillian, Hope and Peter. James is working with Marlan Maritime Technologies to design resilient coastal cities using big data. Cillian is working with Ordnance Survey on improving the Geolocation of emergency service response using big data. Hope is working with the Liverpool City Region researching human dynamics within an urban and regional context. Peter works with Data Fusion Ltd looking at the Geodemographics of British streets. We are part of the Liverpool University Geographic Data Science Lab: a research and teaching centre focussing on making sense of the world by using spatial data to make intelligent decisions.

We have recently hosted our fellow Year 1 students from Leeds, Manchester and Sheffield for a short course which involved using Python for Data wrangling and machine learning, taught by Dr Dani-Arribas-Bel from the GDSL. The course was designed as an introduction to data cleansing, manipulation and visualisation through use of the Python programming language, as well as both supervised and unsupervised machine learning techniques that will no doubt form a fundamental part of our research. Following this week, we have been sent away with an assignment to choose our own datasets and conduct some in-depth data analysis using the techniques we have learned from the course.

Over the last couple of months we have been studying for the Masters component of our CDT programme, as well as developing the ideas and thesis structures of our CDT projects with data partners and supervisors to stand us in good stead for the coming years. We are currently gearing up for a two-week internship with our data partners which will provide us with invaluable industry experience, as well as the opportunity to work with data that we will likely be working on for the next couple of years.

The Liverpool course was a great way to start the New Year and we are looking forward to the Manchester and Sheffield courses coming up in the next couple of months!

CDT students show excellence across the board in 2019 Partner Event

 

The 3rd year student organisers of the second CDT partner event are happy to report on a successful day. The event was hosted again by the Leeds Institute for Data Analytics (LIDA) and while the essence of the day was similar to last year, this year saw the addition of mini masterclasses to the bill. The academic staff and student attendees from the four CDT institutions were again joined by representatives from partner organisations and got a glimpse of the work being undertaken by students.

Professor Mark Birkin, LIDA Director, opened proceedings with a short welcome speech before handing over to University of Liverpool’s Professor Alex Singleton who chaired the day’s first session of individual lightning talks by 3rd year students. These replaced last year’s group presentations and provided a 3 – 4-minute snapshot of the research conducted for master’s dissertations or first papers. These bite sized presentations highlighted the range of research areas and skills being used, from assessing the impact of the weather on high street retail to examining inequalities in cycling participation. Students took questions after their talks and thanks to Dr Mark Taylor from the University of Sheffield asking us for our take home message, we all now have an elevator pitch of our work.

This year’s poster session was taken on by the 2nd year students with feedback highlighting their excellent knowledge and enthusiasm for their work. Examples of work completed for core modules could be seen in posters detailing the use of web scraping and text analysis. (You can view the event posters here!) Again, the variety of topics and analysis methods on show highlights the diverse range of projects undertaken. There truly is something for everybody on the CDT.

The partner event was timed to coincide with the Introduction to Programming module, the first of the MSc, and we were joined at lunch by the new cohort of students. You hit the ground running with this module, particularly if you’re new to coding, so this year’s lunch was extended to include a Q&A session hosted by Dr Eleri Pound (Centre Manager) to cover any questions or queries the new students had. A range of questions were submitted and 2nd and 3rd year students were able to pass on words of advice, encouragement and to hopefully alleviate any concerns.

The day’s final session was the mini masterclass. There was the option to sign up to one of three masterclasses; academic publishing, networking or public engagement. Feedback shows that attendees of each of the classes found them helpful and informative. The event overall was enjoyed by everybody, with comments showing that people found it interesting and engaging. Remember to follow our twitter page, @DataCDT, to stay in touch with CDT students as we continue to work at the cutting edge of our subject areas.

Written by Melanie Green, Noelynn Onah & Rhiannon Thomas.