"I chose to pick an inherently difficult problem"

© 2021 EPFL

© 2021 EPFL

Michael Mark, joined the team of Prof. Thomas Weber at EPFL's College of Management in Technology in 2017 to pursue his studies and obtain his doctorate this year. He tells us about his experience so far.

Can you describe yourself in a few words?

I grew up in Brno, the second largest city in the Czech Republic, where I began my undergraduate in engineering at Brno University of Technology. After my first semester, I transferred to Université Pierre et Marie Curie in Paris where, in 2015, I graduated with a degree in mathematics. By the end of my studies, I was involved in an industry project in quantitative finance which naturally led me to Leicester, United Kingdom, where I earned my Master’s degree in financial mathematics. I then landed a position at Ernst & Young in London as a quantitative risk analyst in one of its advisory divisions. My work at EY solidified my interest in data-driven decision making and, with the support of my engagement manager, I decided to pursue a doctoral degree in operations research. Finally, since 2017 I have been a PhD student in the EDMT program at Operations, Economics and Strategy (OES) chair of Prof. T. A. Weber.

Tell us a bit about your research and how you think it will develop in the future?

My research revolves mainly around a special type of a stochastic point process known as a Hawkes process. The salient feature of a Hawkes process is its self-exciting nature, which makes it extremely well suited for modelling of various phenomena ranging from earthquakes and aftershock sequences to high-frequency trading and financial contagion. In my research, I inspect the process from different angles, including its various theoretical properties or its other, novel applications to already established problems.

Currently, I am working on a learning agent for a general class of controlled Hawkes processes using the popular reinforcement learning paradigm. The application we have in mind is tied with the problem of credit collections; however, our results are readily extendible to other sectors. For instance, our method incorporates expert domain knowledge into the agent’s learning process, which is of interest in numerous practically relevant reinforcements learning applications.

Can you describe your typical day during the semester and during semester break? How have you adapted to the Covid-19 crisis?

During the semester break my entire week is dedicated to my research projects, so on a typical day you would find me either reading papers or coding behind my desk. It is a perfect time to deep dive into a project as there are usually no other distractions. With the semester starting a portion of my time is allocated towards my teaching duties, be it correcting problem sets, holding office hours or teaching exercise sessions.

The Covid-19 outbreak clearly negatively impacted the work of most researchers at EPFL. However, compared to my colleagues in other fields who rely on special equipment and lab access I only require my laptop and internet connection; therefore, I had an easier time to adapt.

Tell us about the atmosphere within the chair of Prof. Thomas Weber and with your fellow researchers?

For the past three years I was the only PhD student of Prof. Weber, which, frankly speaking, was extremely convenient as I could profit from one-on-one meetings almost any time I needed – a luxury most PhD students probably will not experience. Our sessions have always been very hands-on and thanks to the professor’s guidance and feedback I have been staying on track with my research projects.

This year, we have been joined by three additional PhD students and I am looking forward to a joint OES research project. We have several open problems in the Hawkes process line of work so I am excited for the upcoming collaborations.

You plan to graduate in 2021, do you know already where and in which field you would like to pursue your career?

The part that most enthralled me during my PhD was picking an inherently difficult problem and having a go at it. Combined with the other perks that academia has to offer, I am already on a lookout for a suitable academic position. That being said, there are a number of opportunities in the financial industry here in Switzerland that are research related and closely match my profile and interests. Therefore, I keep my options open – I have to defend my thesis first.

Outside of research, do you have another interest?

I played beach volleyball competitively since high school. Back in the Czech Republic I played for the national team under eighteen and twenty and then, when I moved out for my studies, I competed for varsity teams both in the UK and in France. When I first arrived to Switzerland I wanted to carry on with the tradition, however, it soon became clear that I would rather spend my free time outside in the Alps than inside in the gym. Over the winter I spend most of my weekends skiing the snow parks of local resorts or hunting for powder in the back country, while summer is dedicated to climbing and mountain biking. I am already excited about the upcoming and hopefully snow-abundant winter.

Author: Alexandra von Schack