Causal Machine Learning for Reliable Decision-Making
MCML Authors
Dennis Frauen
Dr.
* Former Member
Abstract
Dennis Frauen
Dr.
* Former Member
Abstract
This dissertation develops methods for reliable data-driven decision-making using causal machine learning. It addresses key challenges such as unobserved confounding and missing counterfactuals by focusing on robustness and efficiency. The work introduces methods for sensitivity analysis, partial identification, and novel meta-learners for estimating treatment effects. Together, these contributions enable more trustworthy decision-making in complex real-world settings. (Shortened.)
BibTeXKey: Fra26