Proposes Mutual Information Surprise (MIS) framework and reaction policy using sampling adjustment and process forking that outperforms classical surprise measures on synthetic and pollution estimation tasks.
Dynamic exploration–exploitation trade-off in active learning regression with Bayesian hierarchical modeling,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Mutual Information Surprise: Rethinking Unexpectedness in Autonomous Systems
Proposes Mutual Information Surprise (MIS) framework and reaction policy using sampling adjustment and process forking that outperforms classical surprise measures on synthetic and pollution estimation tasks.