Adaptive Bayes exactly tracks information over intrinsic time

2026-06-26

AI Summary

本文通过引入一种精确的信息追踪方法，解决了贝叶斯更新和乘法权重更新中的累积遗憾问题，并提出了一种基于内在时间的学习率调整策略。

Abstract

Bayesian and multiplicative-weights updates reweight experts, models, or actions from sequential feedback. We show that the regret of any such update obeys an exact information-accounting identity. On each round, the learner's excess loss to any chosen comparator is the sum of an immediate payment for the uncertainty exposed by the round and a reduction in the information distance from the learner's current weights to the comparator. The cumulative payment defines a pathwise uncertainty clock, the \emph{intrinsic time} of the realized sequence. Summing one-step balances yields two exact adaptive decompositions of cumulative regret, one for each natural way of composing the update across rounds. Because the decompositions are exact rather than upper bounds, favorable stochastic or low-noise regimes appear as self-bounding properties of the realized intrinsic time, not as slack in worst-case analyses. The same calculus covers Hedge, optimistic and side-information variants, continuous priors, boosting, online convex optimization, contextual bandits, and repeated games: the pathwise account is the same in every case.

Problem

• Bayesian updates
• multiplicative-weights updates
• regret decomposition
• information distance
• intrinsic time

Innovation

• Adaptive Bayes
• information-accounting identity
• intrinsic time
• cumulative regret decomposition
• learning rate schedule