"We Get a Lot of Freedom and Flexibility in Terms of What We Want to Teach"
Geoffrey Decrouez is an Assistant Professor at the Faculty of Computer Science at HSE. He told The HSE Look about his research and teaching at HSE.
– How did you come to work at HSE?
– Before coming here I was in Australia for ten years. I was studying and doing a postdoc there, and later I was looking for a permanent position. I had been to Moscow for conferences a few times during my postdoctoral fellowship. So I already liked the city, I knew several colleagues who are professors here, and I kept an eye on possible vacancies. Of course, I did other job applications as well, but I must say that I was very interested in this position at HSE. It was an easy choice given that during the last three years I was attracted to the idea of moving to Russia.
– What do you focus on in your research?
– I had different specializations over the years – I started with signal processing, and I am still into it these days. My first postdoc was in probability theory and my second postdoc was in statistics, so now I am more and more interested in statistical application. Together with my colleagues at the faculty we are thinking about collaborating more on some machine learning projects; I also have another French colleague in the Faculty of Economics who is doing research in this sphere. Machine learning isgoing to be my main research focus for the nearest future.
– In terms of collaboration, do you see many opportunities here? Is there any language barrier?
– The language barrier is actually not a problem at all – I speak a bit of Russian and my colleagues speak English. Now there is a new dynamic in our department, we want to enhance collaboration, so we are going to meet more often in the future in order to work on common topics. And I also still have my collaborations in Australia and France and I keep them very active.
– Do you teach?
– I am teaching three courses. The first course is in probability theory and statistics. The second one is a MAGoLego course – it is an elective for first-year Master’s-degree students from different faculties. It focuses on statistics in experimental design – it is an introduction to statistics for non-mathematicians. I really enjoy these classes; it's quite refreshing to speak about statistics to non-statisticians because they look at things in a very different way. There are about 25 students who come from the Faculties of Social Sciences, Humanities, Business and Management. So there is a really broad variety of students and backgrounds. My third subject is about machine learning and is more advanced, designed for first-year Master’s-students in our faculty. I teach these subjects in English and they are brand new to our faculty. I hope there will be more such courses later, so that students can take more classes in English.
– Do students work with you on their term papers and Masters’ theses?
– I have two students working with me now. One of them is a fourth-year bachelor student who is implementing a neural network. With two colleagues we proposed a mathematical model describing the dynamics between neurons. At the moment there are a few theoretical results but we still don't really understand how the network is behaving on a large scale. So this project is looking at the general behaviour of the network depending on different types of connections and different parameters of the model.
The work of another student (Master’s level) is related to the multifractal theory. Basically we are looking at another way of representing time series that demonstrate multifractal behaviour.
– Do international exchange students attend your classes?
– There are no international students in my classes yet. We are a very young faculty. I know that there are plans to build connections with several French universities and bring students here on a regular basis. We had a winter school for prospective students this year that I missed, unfortunately. But I would like to take part in the Summer University if the opportunity arises.
– If you had to sum up your feelings about being at HSE Moscow so far and your plans for now, what are they?
–I have been enjoying my time here so far. When I arrived to Moscow, I participated in the HSE Day in September and met many of my international colleagues there, so I would really recommend it for the newcomers. I like teaching all these courses and meeting students. What is really good about working here is the amount of freedom and flexibility we have in terms of what we want to teach. I have been given some directions, but it was up to me to create three new subjects. We don’t have this sort of autonomy in a similar position at many other universities, so it’s been a great experience.
As for plans, at this stage I am really doing a lot of independent research as I have several ongoing projects that I need to finish, and with all the teaching I am yet to start new collaborations. Right now I am focusing on finishing my ongoing research papers and moving to the new projects with my colleagues at the School of Data Analysis and Artificial Intelligence here at Computer Science Faculty and Quentin Paris, a colleague from the Faculty of Economics.
Geoffrey Decrouez was born in Dunkerque, France. He studied electrical engineering and signal and image processing in France for his undergraduate degrees. In 2009, he defended his PhD dissertation on The Generation of Multifractal Signals with Underlying Branching Structure at the Department of Mathematics and Statistics, the University of Melbourne. After receiving his PhD, Decrouez conducted research in stochastic modeling, with applications in neural networks and finance as a postdoctoral fellow at the University of Melbourne. In September 2014, he started working as an Assistant Professor at the Faculty of Computer Science at HSE. In his free time Decrouez enjoys exploring Moscow, where he has quite a few Russian and international friends.
The full text of the issue can be found in The HSE LooK 5 (22), May 2015. If you are not on our regular mailing list yet, please subscribe and get fresh issues of our bulletin every month!
Geoffrey Gerard Decrouez