Principle of maximum entropy
Principio de máximo de entropía
- Date
- 1957 · confidence high
- Place
- Stanford, California, United States
- Credited to
- Edwin Thompson Jaynes
- Field
- Cosmos & Earth
What was discovered
A principle that, among all probability distributions consistent with known constraints, selects the one that maximizes Shannon entropy, avoiding assumptions not justified by the available information.
What it led to
- became a standard tool of Bayesian inference
- influenced statistical mechanics and information theory
- was applied in fields such as signal processing and machine learning