Joel Persson
Research Scientist
Spotify Research
Hi, my name is Joel (Swedish pronunciation "yo-el") and welcome to my website.
I am a Research Scientist in the Machine Learning, Economics, and Optimization lab at Spotify, where I work on methods for heterogeneous treatment effects, contextual bandits, and off-policy evaluation to improve personalization and experimentation. More broadly, I am interested in causal inference, machine learning, and data-driven decision-making, in particular for problems in digital domains such as digital marketing, platforms and health.
I hold a PhD from the Department of Management, Technology, and Economics at ETH Zurich, supervised by Stefan Feuerriegel and Florian von Wangenheim. During my PhD, I visited the Operations, Information & Technology area at Stanford Graduate School of Business, hosted by Jann Spiess, interned as Machine Learning Research Scientist at Booking.com in Amsterdam, and joined the non-profit startup Algorithm Audit as a core technical contributor. Previously, I obtained double BSc and MSc degrees in Statistics and Business & Economics from Lund University, Sweden, and worked on marketing science problems at GfK Nielsen and in digital advertising at an award-winning European agency.
Research interests:
Causal Inference
Statistical Machine Learning
Data-Driven Decision-Making
Personalization and Digitalization