I am a 27-year-old German Data Science and Artificial Intelligence PhD student at the University of Edinburgh.
After graduating from Saarland University in August 2018 with my Bachelor’s degree in computer science, I moved to the beautiful city of Edinburgh, Scotland, to continue my studies as an Informatics postgraduate student. Following my graduation with Distinction and my M.Sc. in Edinburgh, I started a PhD to research novel exploration and generalisation techniques for multi-agent reinforcement learning as part of the Autonomous Agents Research Group.
Computers are omnipresent in our modern society. We use them on various occasions during our everyday life, often without even noticing. While smartphones, personal computers and laptops are clearly recognised as computers, it is less visible that most cars, washing machines, microwaves, TVs and generally most electronic devices contain some form of a simple computer. It is quite astonishing that these tiny chips are often more capable than the first programmable, electromechanical computers in the 1940s which filled entire rooms like the Zuse Z3.
I always found this flexibility and computational power of modern computers impressive. They are used to solve a variety of problems and despite, at an earlier age, hardly knowing anything about how they operate, this alone raised a curiosity in me. I did not just want to use computers on a daily basis but understand how they work.
This led to me choosing computer science as a subject for the last three years of my school career. There, I was introduced to programming and a variety of concepts used in computer science, which I would later learn about in more depth during my studies at university. These lessons strengthened my interest which led to me studying computer science at Saarland University after finished my Abitur (German high school equivalent).
At university I quickly noticed that despite experiencing computer science for three and mathematics for twelve years at school, this hardly scratched the surface of what these fields contain. While the first semester surely was an immense challenge, this deeper dive into math and computer science required entirely new ways of thinking. Applying mathematical concepts for formal proofs requires a significantly deeper understanding than simple calculations which dominated mathematics at school. The first two years of my undergraduate studies mostly involved a variety of foundational courses covering objective, imperative as well as functional programming, concurrency, data structures and algorithms, theoretical computer science, mathematics with analysis, linear algebra, statistics and more. Afterwards, I was specifically interested in artificial intelligence. The idea to express knowledge, logical thinking and learning with the help of mathematical models is immensely interesting to me.
The AI courses I attended covered mostly two areas. For one, I learned about foundational concepts of machine learning and applied this knowledge in a practical course focusing on neural networks and deep learning as a part of machine learning itself. Not just learning about these in theory, but also directly applying and implementing these models significantly deepened the understanding in this subfield of AI.
Secondly, I attended multiple courses regarding automated planning. The big goal of this research field is to implement a single planner which is then able to solve a variety of problems. The flexibility of these planners fascinated me and therefore I also wrote my thesis "Domaing-Dependent Policy Learning using Neural Networks in Classical Planning" aiming to combine deep learning and automated planning.
Following my Bachelor’s degree, I decided to continue my studies as an Informatics postgraduate student at the University of Edinburgh. Studying in Edinburgh was a truly remarkable experience. I grew not just as a computer scientist and researcher but also as an individual gaining life experience living and studying abroad in this international environment of academic excellence.
My M.Sc. studies mostly focused on machine learning and robotics, taking courses covering pattern recognition, theoretical machine learning, probabilistic modelling and reasoning. Lastly, I studied the areas of game theory, decision making and reinforcement learning with especially the last field fascinating me immensely. Hence, I dove deeper after an introductory course in my M.Sc. dissertation covering Curiosity in Multi-Agent Reinforcement Learning in which I analysed the impact of intrinsic curiosity for exploration in competitive and cooperative multi-agent reinforcement learning environments.
Following this research, I started as a Data Science and Artificial Intelligence PhD student at the Autonomous Agents Research Group in December 2019. Since then, I managed to work on my first publications together with my excellent colleagues at the research group in Edinburgh and visited my first research conferences, NeurIPS 2020, AAMAS 2021 and ICML 2021, as a presenting researcher. While I am still waiting for the first in-person conference experience, I am also looking forward to NeurIPS 2021 where we will present a MARL benchmark paper. Besides my academic studies, I also managed to gain invaluable research experience applying my expertise in an internship working with Dematic on warehouse automation. I really enjoyed this experience which showed me a different perspective on research. I am looking forward to the future and am hoping to make a lasting impact in multi-agent reinforcement learning.
Travelling is highly rewarding. It allows us to broaden our horizon in many regards. We can see beyond our doorstep, explore the beauty of planet Earth and experience a variety of cultures and people we could not meet in our usual daily life. While these encounters and experiences can seem daunting at first, I encourage everyone to dare, you won’t be disappointed.
It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.
I have to say, I was fortunate to grow up with parents fond of travelling themselves and who took me and my brothers along many fantastic journeys. However, the most influence experiences for me were my trips to Japan in 2017 and 2019 when I travelled just with friends to Tokyo, Osaka, Kyoto and Hiroshima. Planning and dealing with everything that comes with such a journey gave me an entirely new appreciation for travelling. Above are some impressions of travel experiences I had in the past years.
I read (almost) everyday for many years alrady. And if I say almost, I refer to the few exceptions in the year I can count on the fingers of one hand. It became a habit that I read just before sleeping, so that now I even have trouble falling asleep if I don’t. In this regard, reading became part of my daily routine. However, it is much more than just a habit. Reading allows me to experience fantastic worlds, heartbreaking stories and learn anything imaginable. It can be inspiring, touching and exciting while all of it is taking place in our imagination.
For many years, I read almost exclusively fantasy novels. From Harry Potter, which I still cherish to this day, over Eragon, the Dwarves and many more to Lord of the Rings. I just could not get enough of stories about magic, dragons, elves and all these fantastic creatures I would only ever be able to meet in my imagination. However, after quite a few years of exclusively reading fantasy I started to become bored of it.
After a while, I dared to read something entirely different and rediscovered my passion for reading by reading historic novels of Ken Follett, motivated by my father who shares my passion for history. From there I started reading a larger variety of books from science-fiction thrillers like The Swarm of Frank Schätzing, the biography of Steve Jobs to Life Without Limits. In this truly inspiring and touching book, which I can not recommend enough, Nick Vujicic shares his story of becoming a motivational speaker being born without any limbs.
Besides the impressions one can get from reading, it can also help to improve language skills. I surely became more proficient in English from reading many novels in this language. I would highly encourage everyone to read frequently. No matter your interests, there will be books matching your preferences and it is rewarding, beneficial and fun at the same time.
My taste in music is rather broad. I listen to a large variety from classic piano music, orchestra (particularly movie soundtracks), acoustic and pop music to hard, punk and old-school rock. To name a few of my favourite artists: Ludovico Einaudi, Hans Zimmer, Howard Shore, Ed Sheeran, One Ok Rock, Greta Van Fleet and Radwimps.
Besides listening to music, I started playing the guitar at the age of 7 and had weekly lessons for 11 years. While I don’t play as frequently anymore as I used to, I still try to maintain parts of what I was able to play. Making music for me was always an outlet whenever I am stressed. Fairly recently, inspired by my younger brother and mother, I also started to learn playing the piano. I am not aiming to become proficient at it but I enjoy the flexibility of the instrument.
Below are some samples I recorded using just a acoustic guitar and a TC Electronic Ditto X2 Looper:
Being born in 1996 I grew up with Gameboys, Pokemon and Zelda. This first introduction to gaming hooked first my older brother and briefly after also me and then my younger brother. While the games we played became increasingly advanced and complicated over time, playing on playstation consoles, Wii, Gamecube and later computers, our passion for gaming remained.
Similar to books, I was mostly drawn to fantasy and roleplaying games at first, but also enjoyed more competitive real-time strategy (RTS) games like The Lord of the Rings: The Battle for Middle-earth II and Starcraft II, which I played against my brothers or the in-game AI. This way, I was first introduced to AI in the field of video games. While, at the time I was already astonished that a computer-controlled player could be much better than myself, I did not think much about it.
With more complex, competitive games I also started to play League of Legends. I never became really competitive, aiming to become better and better, but still enjoyed the complexity and learning aspect of the game, which certainly can be a lot of fun, particularly with friends. Through this game, I started following the professional Esports scene, which boomed in the last years. While I nowadays find fewer time to play myself, I am excited and fascinated by the level of play and strategy involved in competitive gaming.
Throughout my Masters studies in Edinburgh, I became increasingly interested in artificial intelligence as a whole and particularly machine learning. I discovered reinforcement learning and its results in video-gaming reaching superhuman performance in Atari game playing. The field is quickly moving, driven forward by research of large industry-based research-labs like Deepmind and OpenAI as well as various academic researchers. As I am passionate about recent research done in the field, I hope I will have the opportunity to contribute in the future.
