Hey! I’m Moritz, a PhD student at the Institute for Neural Computation (INI) at Ruhr University Bochum.
My current research is focused on representation learning for reinforcement learning. With the right representations of input data, reinforcement learning can become not only more powerful but also more explainable and thus transparent. I believe this will eventually help to transfer reinforcement learning from academia to real-world applications.
If you’ve got any questions or suggestions don’t be afraid to drop me an email!
About this blog
This blog is intended as a way to share my discoveries and research results, in a boiled-down and less rigorous way than peer-reviewed publications would allow for. I hope this makes them accessible to a larger audience of students and fellow researchers.
A bit of history
I started this blog as a way to share the knowledge I gathered when I started reading about representation learning. As of 2021, there were - in my opinion - no good, comprehensive sources explaining what representation learning in machine learning is, or how it can be used. Instead, I only found myriads of papers casually using the term or proposing methods that apparently do representation learning, but without ever defining or reviewing it.
The very first series of blog posts on representation learning was intended to help other researchers get into the topic without having to sift through all those papers I ended up reading.