marten

Research Advice for Myself

It has not even been half a year since I started my second foray into research and this time I want to do a few things differently. One of them is to follow advice from more experienced researchers such as my advisor without questioning it too much. I trust that their approach is more informed than mine and that I will progress faster as a result. A recent example is the importance of fast prototyping over first gaining a broad understanding of a method or field. I can see the reason and yet it goes against my intuition.

In addition to people in my group I also have access to advice from people all over the world thanks to the internet. Here, I will collect my various bookmarks and list the points that resonate the most with me.

marlow41 listed some habits on reddit that they would adopt from the beginning if they were to begin their PhD today which mostly agrees with Sara Billey. First, research should always take priority over other activities. You might have other duties such as teaching or talks but you should try to make some actual research happen every day. Their tool of choice is a journal to keep a list of accomplishments but also a strict table of research time to keep yourself accountable. Second, you should always be reading a paper. The goal is to make yourself aware of methods that you might need in the future. Finally, you have to advertize your work. Give talks, talk to people, be active on the internet.

Michael Nielsen and John Schulman both have written extensive essays on researching effectively in the field of machine learning. My main takeaways from Nielsen's text are the importance of developing a research strength, a field of expertise, and how he sees the creative process as two activities, the problem solving and the problem creation. To me it seems useful to be aware that progress can come in different forms that rely on different skills.

Schulman gives another split of research into idea-driven and goal-driven, orthogonal to Nielsen's. The former roughly focusses on improving existing methods with new ideas whereas the latter is formulated in terms of a, usually applied, goal such as make X work for the first time. He also emphasizes both the development of a "taste" for problems and keeping a journal, in line with the other authors.

Mathematician Terry Tao has lots of advice on writing some of which even generalizes across the boundaries of mathematics. In particular, I want to follow his advice on prototyping a paper. I am convinced that having a prototype paper from the beginning helps drive your thought process and it also agrees with the number 1 advice from Prof. SPJ.

Recently, I have begun to prototype more quickly and keep a hand-written journal of observations and thoughts. In the future, I need to find a way to make consistent progress on my research while also keeping up with teaching. Regarding research direction, I feel an inherent desire to do idea-driven work. It is pure and beautiful. At the same time I believe that my profile and skills are much better suited to goal-driven research. So I will also need to figure that out.

Want to comment or get in touch? @martenlienen or email me