background

I'm originally from Moscow. I studied at the Moscow Institute of Physics and Technology, Department of Control and Applied Mathematics. I had a classic MIPT curriculum in physics and a series of various mathematical disciplines: mathematical analysis, linear algebra, mathematical statistics, probability theory, theory of random processes, etc. In my fourth year of studies, I found an internship in Samsung R&D in a data analysis team, where I worked on smartphone sales prediction. Since then, I got interested in Machine Learning and decided to get more profound knowledge in the field. So, in spring 2019 I joined Skoltech's Data Science (MLAI track) program.

After my Bachelor's studies were over, my goal was to deepen my knowledge in programming and machine learning. I knew that Skoltech had one of the best studying programs in machine learning in Russia. I also appreciated the education system at Skoltech: having only several compulsory courses and being able to choose the rest as I wished. Not to mention the stipend (which equals a common salary for a 40-hour internship in an IT company). Thus, the choice of Master's program was obvious.

The Data Science program covers different themes and aspects of ML and DL. It starts with basic courses of algorithms and numerical linear algebra as well as introductory courses in data science and computer vision. The program continues with classic ML and DL courses and may be followed by more specific courses such as Introduction to Blockchain. This educational path gives a broad and structured view of the Data Science field.

When the time came to choose my research at Skoltech, the projects of the Joint Laboratory of Applied Research Skoltech-Sber (LARSS) seemed attractive to me. I also liked the idea of being able to work in a major company (Sber) while studying. A great advantage of this particular program is that it allows combining work and studies smoothly without interfering with one another, which is less likely for other programs.

In LARSS, I worked on several projects, including the interest rate elasticity of demand for loans, ATMs loss prediction for the next month, transactions analysis to identify car insurance (KASKO mainly) buyers to be able to make special offers and many others.

Looking back at this experience and this unique internship opportunity with Sber, I definitely feel like I've gained enough hands-on skills in machine learning to become a sought-after expert after graduation. Moreover, after finishing my project in the lab, I stayed on as an official Sber employee to implement my project into production and explore new areas.

My diploma is one of the projects I completed at Sber. It is titled "Interest rate elasticity of demand for loans" and is about measuring the elasticity of demand. Various clients react to the changes in the interest rate suggested to them quite differently. My task was to assign a corresponding value of sensitivity to the interest rate change for every client. It is of high interest for a bank as it allows to increase the share of loans taken by identifying sensitive clients and lowering the rates for them. The novelty of the method is that it works like a targeted advertisement. It makes it possible to identify those clients who react to lowering a rate the most.

It's exciting to see my research actually being applied at Sber. Now I'm working on my diploma project, trying to implement different methods which go beyond the bank's needs, as well as other Sber projects. One of them is connected with real estate price prediction to be used in the process of risk calculation when issuing mortgages.

I would like to build a career in the Machine Learning industry. The research work in LARSS opens a lot of doors and provides an ability to see how ML can be applied in a financial sector. I would like to continue growing as an expert in Data Science and become a Team Lead eventually, as I enjoy both technical and managing and communicational aspects of work. Working in an industry is a great opportunity to learn how modern economy evolves and to be able to influence its future