PhD position: machine learning for human learning

Dear Julians, transformers of code, wielders of symbols and squeezers of eps,

We are offering a PhD position at the ML ⇌ Science Colaboratory in Tübingen, Germany.

We seek to develop an adaptive learning approach that works on structured environments - domains where the learning order of concepts is very important. Programming, including machine learning, databases and an API is a very important component of the project, and we’re friendly to Julia. We aim at real-world applicability: this free-software backend should be useable by many frontends (including Anki, emacs, Tiddlywiki…). Ultimately we aim at making self-directed learning more enjoyable and effective for everyone. We would like to combine the ability we have now to infer knowledge graphs from text, with what we know about how brains and machines learn, with good software engineering.

Some keywords: spaced repetition, graph neural networks, Gaussian processes on graphs, reinforcement learning, graph databases, REST APIs, psychophysical experiments with human learners, …

Learn more about the position here:

We are empthatically inviting people who are underrepresented in science, technology, computing and mathematics to help us become the diverse, balanced group we strive to be. Apply, ideally until the 30th of April!


Our deadline is approaching!

If you are an avid learner and you would like to have a machine learning algorithm propose what next concept you could learn based on what you know (and remember), where you want to go, and how the domain itself is structured, then come figure it out with us!

The intellectual environment in Tübingen is very stimulating with the top ML community in Germany, and the project is part of a multidisciplinary effort to apply machine learning to education, with supervision from multiple fields.

If you are interested, write to me for questions or apply directly,