{"id":"arxiv/2605.02099","url":"https://arxiv.org/abs/2605.02099","tags":["math.ST"],"score":0,"title":"Entropic Strict Minimum Message Length and Its Connections to PAC-Bayes and NML","domain":"math.ST","source":"arxiv","authors":["Enes Makalic","Daniel F. Schmidt"],"reviews":0,"subject":"math.ST","summary":"We introduce entropic strict minimum message length (SMML), a risk-sensitive generalization of strict minimum message length coding. The proposed criterion replaces expected two-part codelength under the prior predictive distribution with an exponential certainty equivalent, thereby defining a one-parameter family of coding rules that interpolates between Bayesian average-case coding and worst-case minimax coding. We show that ordinary SMML is recovered in the risk-neutral limit, while the extreme risk-sensitive limit yields a minimax codelength criterion. Applying the same entropic soft…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-05-03","reactions":{}}