Groups
Category
Level
Minimum Description Length (MDL) picks the model that compresses the data best by minimizing L(M) + L(D|M).
Rényi entropy generalizes Shannon entropy by measuring uncertainty with a tunable emphasis on common versus rare outcomes.