Welcome!

I’m Andreas. This is my little hypertext garden on the Internet — make yourself at home!

Currently, I’m figuring out how best to incorporate biological and geometrical priors for representation learning; that’s the working title of my PhD! To what degree can we utilize expert knowledge for guiding the learning process? Does it help? These types of questions are at the heart of my research!

I’ve been put in the deep water of biological machine learning by my PhD supervisor Anders Krogh who I am grateful to be working with. After he and Viktoria Schüster created the Deep Generative Decoder, a simplified VAE-like model with no encoder, I’ve explored various ways of getting the most from decoders. Luckily, geometrical machine learning has already analyzed such settings to a great extent, providing a great starting ground.

On a broader level, I care a lot about learning and playing. When I’m not in the office working on my PhD, you’ll probably find me struggling with Chopin, competing in games, learning languages, or just exploring the world — most often through film and animation ;)

As far as ‘about’ pages go, I hope this acts as an open invitation into this new garden. You may be interested in:

• The blog desk: a layman-oriented version of my academic work, and the raison d’être of this garden;

• The visual curriculum vitae: my educational and professional path in a super concise format.

Andreas Bjerregaard :)