Mathias Drton
Mathias Drton studies graphical models and statistical inference. His research asks when observed data can support conclusions about causal relationships. Drton earned a statistics doctorate at Washington in 2004 and later worked as a postdoctoral researcher at Berkeley. He began his Chicago faculty career in 2005. He became a statistics professor at Washington in 2012, spent a year at Copenhagen in 2018–2019 and became a mathematical-statistics professor at TUM in 2019. Drton coauthored the half-trek study with Rina Foygel Barber, Nils Sturma and Luca Weihs. In specified linear structural equation models with latent variables, their sufficient criterion identifies direct causal effects between observed variables from the observed covariance matrix.
- Works
- 197
- Citations
- 4,707
- h-index
- 34
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
2004–2005
Doctoral and postdoctoral research
Drton earned a statistics doctorate at Washington in 2004 and later worked as a postdoctoral researcher at Berkeley. He began his Chicago faculty career in 2005.
-
2012–2019
Statistical research across universities
He became a statistics professor at Washington in 2012, spent a year at Copenhagen in 2018–2019 and became a mathematical-statistics professor at TUM in 2019.
-
2022
Identifying direct causal effects
Drton coauthored the half-trek study with Rina Foygel Barber, Nils Sturma and Luca Weihs. In specified linear structural equation models with latent variables, their sufficient criterion identifies direct causal effects between observed variables from the observed covariance matrix.
-
As of October 2026
Algebra and causal inference
Drton uses algebra and probability theory to study which data and model conditions allow causal conclusions. He also seeks efficient methods to estimate those relationships.
Sources and credits
Photo: ellis.eu; Mathias Drton profile photograph; creator not stated. License: Not stated by source. Source: ellis.eu. Cropped to a square and resized. Ask us to remove or replace this photo.