Alexandra Carpentier: there’s magic in research
After beginning her career in France and then going to Britain, the Franco-Greek researcher Alexandra Carpentier is preparing to move to Germany: a journey marked by her desire to understand the fascinating universe of mathematical statistics and machine learning.
What is your field of research?
I’m interested in statistics, particularly statistical learning and machine learning. Basically it’s about modeling ways of learning and making decisions to automate optimum strategies for solving tasks. To achieve this, you can take inspiration from the decisions made by biological organisms to survive or you can try to understand the mechanisms of human learning. It’s just fascinating. Statistics and machine learning are lively research fields with a young, friendly community.
These were issues that you explored at Inria, weren’t they?
After graduating from engineering college I joined the Sequel project team (a joint venture between Lille 3 and Lille 1 universities*) at Inria Lille – Nord Europe in 2009 to write my thesis. The team specializes in machine learning and among other things I worked on brain-machine interfacing problems with neuroscientists in the Athena project team at Sophia Antipolis – Méditerranée. Together we used active learning techniques to improve these interfaces. I have very fond memories of the three years I spent in Sequel: I developed a real passion for research there. I particularly enjoyed being in an environment where mathematicians and software engineers mixed together; they complement each other really well. I also made some very good friends among both the researchers and the administrative staff.
Why did you decide to leave Lille and go to Cambridge?
To follow my partner! But it was also a great career opportunity as I was able to join the Stats Lab in Cambridge and get stuck into problems that are a bit more statistical. As a research associate I am employed directly by the university, so I’m fairly free in terms of research areas. This is really beneficial for young researchers, who are often dependent on funding.
How would you sum up your experience in Britain?
From the scientific viewpoint, it’s hard to generalize as the Stats Lab is a bit of an exception in the UK and like France statistical research is quite theoretical in Britain, which may be why eight out of the forty team members are French! Anyway, scientifically and personally my experience has been very rewarding and enjoyable and I have learned a lot.
From the cultural viewpoint, it was a bit like being plunged into the world of Harry Potter: it’s not unusual to see people walking round in cloaks in Cambridge and I also have the opportunity of attending formal dinners in the colleges, where the professors sit at a top table and the students sit below. In traditional colleges the meal starts with a prayer in Latin. On the other hand, I was surprised to hear everyone being on such familiar terms, even signing themselves with just their first names on semi-official documents! That would be unthinkable in France.
But now you’re moving to Germany in September
Again I’m accompanying my partner: she’s German and wants to go back to live in her country! More seriously, it’s another excellent scientific opportunity for me because as soon as I get there I will be receiving a five-year research grant** from Potsdam university. I’m planning to form my own team to study statistical anomaly detection and develop practical applications. I’m looking forward to working closely with the business world, as happens a lot at Inria.
How would you sum up the world of research in the light of your initial experiences?
I’ve noticed that there are not many women in mathematical research and that’s a pity. Even though things are changing slowly, it’s sometimes hard to be under-represented, especially when you’re making progress. In terms of funding, there are tremendous disparities between countries: lectureships don’t exist in Germany but there are personal grants for young researchers who want to put a team together, like the ones provided by the DFG, and I think this is a really good system.
** within UMR 9189 CNRS-Centrale Lille-University Lille1, CRIStAL
** The Emmy Noether grant is awarded by theDFG(Deutsche Forschungsgemeinschaft), the equivalent of theANR(Agence Nationale de la Recherche) in France.
Alexandra Carpentier, 27, is currently a post-doctoral researcher at the Stats Lab in Cambridge. She began her professional career in 2010, when she joined the Sequel project team at Inria Lille – Nord Europe, where she worked on statistical learning and understanding the human brain.
Between equations, she particularly enjoys crunchingSpeculoscookies!
2011: Publication of her first scientific article on bandit algorithms: simple modeling of artificial intelligence and learning problems.
2012: Completion of her thesis.
2015: Move to Germany and creation of her own research team.