Minimum Description Length
A model selection criterion developed by Rissanen that seeks an optimal trade-off between fitting the data and model complexity, translating Solomonoff's idea of compression as understanding into an operational method.
Timeline (last 90 days)
Solomonoff inductive inference and minimum description length theory point out that the optimal predictor is the optimal compressor.
The Minimum Description Length (MDL) principle is used to evaluate the compression value of a conjecture — if a new theorem can unify the explanation of multiple known results and thereby shorten the description length of the knowledge base, it is considered to have high discovery value
All Facts (2)
Solomonoff inductive inference and minimum description length theory point out that the optimal predictor is the optimal compressor.
50%UnverifiedThe Minimum Description Length (MDL) principle is used to evaluate the compression value of a conjecture — if a new theorem can unify the explanation of multiple known results and thereby shorten the description length of the knowledge base, it is considered to have high discovery value
50%