Interview with ISAM Supervisor Felix Brandt
Winner of the TUM Supervisory Award 2021
The TUM Graduate Council awards this prize on behalf of all doctoral candidates to a distinguished TUM doctoral supervisor who has committed herself or himself in the past to an outstanding supervision and education of her or his doctoral candidates. The TUM Supervisory Award is given on a yearly basis under the patronage of the President of TUM, Prof. Dr. Thomas Hofmann. It is funded by the Bund der Freunde der Technischen Universität München e.V., and encompasses a prize of 5,000 Euros.
This year’s winner, Felix Brandt is Professor in the Group of Decision Sciences & Systems (DSS) at the Department of Informatics and a supervisor for mathematicians. Since he was voted by ISAM members, the ISAM Doctoral Representative Board asked him about his style of supervision and what he considers important when supervising students.
1. What do you consider to be the most important aspects of the role of a supervisor?
In my view, one of the most important tasks of a supervisor is to spark the students’ enthusiasm and passion for research. I don’t think it’s possible to have a successful career in academia if one doesn’t enjoy and is fully committed to research. Of course, not all aspects of academic work (reading, thinking, writing, presenting, serving the community, and teaching) are equally appealing to everybody. It’s perfectly fine to have strengths and weaknesses, but one should always be willing to improve on one’s weaknesses. This is crucial because most of the above-mentioned aspects are interrelated and important to be successful. For example, most young students are eager to prove new theorems, but not so excited about reading dozens of papers. Many problems become much easier to solve after studying the existing literature. An essential skill that aspiring researchers need to develop is to quickly find and interpret relevant information.
One aspect that I personally emphasize a lot is paper writing. Even the most groundbreaking results will not be appreciated if they are written up in a suboptimal or incomprehensible way. Paper writing includes coming up with a convincing storyline, composing a compelling introduction, choosing the right definitions, working on intuitive notation, and going through many rounds of polishing. To me, good papers - just like good movies or good novels - can be enjoyed over and over again. I try to motivate my students by asking them to write the papers they would like to read and, ideally, be proud of ten years from now.
2. How do you make it easier for your students to start their doctoral studies? How do you help them deal with roadblocks in their research?
We have a relatively sophisticated technical infrastructure that (hopefully) helps students become accustomed to our workflow. This includes a LaTeX git repository with all our papers and paper drafts, a huge common BibTeX file with thousands of annotated PDFs, a cloud server with an extensive list of open problems, slides, and posters, and a group guide with guidelines, advice, technical manuals, etc. In particular the BibTeX file is very useful because the papers are organized via keywords and smart lists and all our annotations are shared with each other.
Still, all of this is overwhelming in the beginning, and the thought of being asked to contribute to the huge body of existing research can be quite intimidating. I try my best to help students get over this, but a short initial period of confusion seems unavoidable. I guess this an important aspect of the job: you suddenly find yourself in an office at the university, but in contrast to almost any other job, there is nobody telling you what to do on a given day. The huge amount of freedom—perhaps the greatest benefit of university research—can also be a burden. More so than in any other profession, one needs to be self-motivated. Sometimes, weeks can pass without making any progress. I think a good strategy to counter this frustration is to have a good mix of little and big research questions. Letting a problem rest for a while before attacking it again with newly acquired tools can be helpful too.
3. What is your publication strategy for your students?
Since we work at the intersection of several areas, we can benefit from the advantages of the publishing cultures in different disciplines. Some of our work is submitted to computer science conferences, which have the big advantage of strict submission deadlines. Again and again, I am amazed how quickly you can make progress on a paper when a deadline is coming up in a week. On the other hand, conference proceedings have severe space restrictions and the referee reports can sometimes be superficial. We therefore also publish papers in economics journals, which allows us to spend several years polishing certain papers before they are published.
For students, I think it’s essential to experience the entire publication process numerous times during their PhD (conceiving, writing, submitting, getting feedback, enduring rejection, improving, and eventually enjoying acceptance). In the end, papers are the ultimate “products” of our research. Throughout one’s career, a researcher will most often be judged by his publications (hopefully not in terms of quantity).
Viewing papers as milestones also helps to give structure to the whole process of doctoral studies. I usually don’t have a clearly defined research agenda for new students. This agenda naturally evolves with each additional publication. In the end, we are both surprised how the path turned out to be and how well things fit together in the dissertation.
4. How do you promote international collaborations? How can you help your students build a scientific network?
Since we work in a relatively small international community, this is relatively straightforward. Most of our peers who work on similar problems are not from Germany. International collaborations evolve quite naturally. The conference system in computer science mentioned above also helps a lot. Unfortunately, during the Corona pandemic, online conferences were much less suitable to get to meet new people and establish a scientific network. Thankfully, things are slowly getting back to normal. I am very happy that four of my students will be traveling to an on-site event (their first!) in France next month. These events are important not only for networking, but also for meeting other PhD students and getting to know the people behind significant results.
5. What measures do you take to create a good atmosphere in your working group?
I think it’s essential to have an atmosphere where making mistakes is ok. I try to set an example by making mistakes all the time :). It would be very detrimental for creative work if somebody does not dare to propose a new idea in fear of being wrong.
In general, I am lucky that the size of my research group allows me to have daily group lunches and spontaneous meetings whenever something interesting comes up. This certainly helps a lot in maintaining a positive atmosphere.
Another important aspect is the composition of the group. I am very selective when recruiting new group members. So far, I have been very happy with the performance of each of my students, and they always got along very well with each other and me. I am really grateful to have the opportunity to work with such talented and remarkable individuals.