Master of Data Science Student's Work Praised at Lomonosov Conference

The research of Tatiana Stankevich, a student of the Master of Data Science programme, was voted best in its field at the Lomonosov Conference. Tatiana told us about her participation in the conference and what a “zombie fire” is.

My maxim: to move forward, to develop is to live. This encourages me to try new things, to look for unconventional approaches and solutions, to think out of the box. For this reason, I decided to develop in the field of data analysis. I had encountered data analysis in the course of my education before, but I did not have deep and fundamental knowledge. I keenly felt the need to form a complete picture when I started to pursue scientific research seriously. I also wanted to go deeper into practical work and learn what a data analyst does. I chose the Faculty of Computer Science through word of mouth - I have friends who are studying at HSE University.

In the high school where I studied, a lot of attention was given to preparing students for Olympiads and science fairs. I did my first research at school and reported it at the science fair. Unfortunately, I did not win but I gained experience, which is also important. Later, studying at the university, I took part in conferences and research, and published articles.

Currently, I am working on safety issues at facilities and territories, mainly investigating fire safety which is highly relevant. I am interested in fire risk assessment because predicting the occurrence of fire means saved lives and material values! As firefighters say, "the best fire is the one that didn't happen".

Tatiana Stankevich

About two years ago a friend of mine took part in the Lomonosov Conference and became a prize winner. This inspired me to give it a try. Choosing a section was quite difficult: there were many of them, and each had a variety of sub-sections. In the end, I chose "State and Municipal Management" with the sub-section "Big Data and Artificial Intelligence in State and Corporate Management".

My presentation was devoted to the specifics of applying machine learning to forest fire risk assessment, namely "zombie fires" that occur in Arctic peatlands and can burn for years, smouldering in winter and going into an active phase when temperatures rise in spring. "Zombie fires" in the Arctic zone of Russia deserve close attention at the state level which is currently not paid to them. In order to form optimal management decisions in the field of fire safety, I proposed to develop a system for assessing the natural fire danger of an area, using machine learning algorithms in the process.

I hoped to be in the top three. I was a bit nervous because I was used to presenting in person, not online. I was afraid that I wouldn't manage to present within timeframes or that there would be some network or computer problems during the online presentation, but in the end, everything went without any hitches. I was especially pleased that the report was well received by the jury and I was asked questions and received advice for further work. When the results were announced, I was, of course, delighted: it's nice to show the work and get a high grade. Besides, it was nice to please my supervisor A. A. Garazha that our work won a prize.

This conference gave me a very interesting experience of remote participation. It was organised at a high level, with a qualified organizing committee that quickly solved various problems. I hope that next year (the conference is annual) I will be able, as the organisers have already invited me, to participate again and demonstrate how my research is developing.