The PC algorithm working out a causal graph. It starts with every variable connected to every other, and removes an edge whenever two variables are independent given some of the others. Edges left without an arrowhead are the ones whose direction can't be told from independences alone. Spirtes & Glymour, “An algorithm for fast recovery of sparse causal graphs,” Social Science Computer Review, 1991.
Rickard Karlsson
About me
I'm a scientist and engineer, currently a founding member of technical staff at a yet-to-be-revealed startup. Before that, after learning about algorithms like the one shown at the top of this page, I decided to do a PhD on machine learning and statistical methods for estimating causal effects from data. I'm also a relentless optimist and care about having a positive impact on the world.
Previously
- PhD in Computer Science (cum laude), TU Delft2021–2026
- Machine Learning Scientist, Booking.com2024
- Visiting Research Scholar, Harvard2023–2024
- MSc Engineering Mathematics, BSc Engineering Physics, Chalmers2016–2021
- Data Analyst, NASA Goddard Space Flight Center2019