Maximum Entropy and the Glasses You Are Looking Through
Peter D. Grunwald

TL;DR
This paper interprets the Maximum Entropy principle through game theory, offering a formal distinction in its application to address criticism about its representation dependence, thereby improving its robustness.
Contribution
It provides a game-theoretic interpretation of MaxEnt and introduces a formal distinction in its application to mitigate representation dependence issues.
Findings
MaxEnt can be understood via game-theoretic frameworks
The formal distinction reduces representation dependence
Improves robustness of MaxEnt applications
Abstract
We give an interpretation of the Maximum Entropy (MaxEnt) Principle in game-theoretic terms. Based on this interpretation, we make a formal distinction between different ways of {em applying/} Maximum Entropy distributions. MaxEnt has frequently been criticized on the grounds that it leads to highly representation dependent results. Our distinction allows us to avoid this problem in many cases.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsStatistical Mechanics and Entropy · Bayesian Modeling and Causal Inference · Complex Systems and Time Series Analysis
