Justin Domke
Papers
1
Total Citations
6
H-Index
1
About
Justin Domke is a leading researcher in machine learning, with a primary focus on Bayesian inference, probabilistic graphical models, and decision-making under uncertainty. His most influential work centers on developing principled frameworks for robust action selection, where he bridges the gap between probabilistic reasoning and real-world decision costs. In his notable paper "Loss-Calibrated Monte Carlo Action Selection" (2015, 6 citations), Domke introduced a Bayesian decision-theoretic approach that enables systems—from plant control to robotics—to hedge actions against state uncertainty, thereby minimizing the risk of low-probability but catastrophic outcomes. This work exemplifies his broader contribution: creating computationally tractable methods that integrate loss functions directly into inference, ensuring that decisions are not just probabilistically sound but also practically safe. Domke’s research has been instrumental in advancing the field of approximate inference, and his clear, rigorous exposition has made complex topics accessible to students and practitioners alike. His ongoing work continues to shape how autonomous systems reason about risk, making his contributions essential reading for anyone interested in the intersection of probabilistic modeling and decision theory.
Research Focus
Key Achievements
Top Papers
- 1Loss-Calibrated Monte Carlo Action Selection6 citations · 2015