Necessary and sufficient conditions for realizable Bayes-consistency under general metric losses are the absence of infinite non-decreasing (γ_k)-Littlestone trees with γ_k → ∞.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.LG 2verdicts
UNVERDICTED 2representative citing papers
First strongly Bayes-consistent algorithm for metric-valued regression with unbounded loss in the agnostic setting, based on metric medoids and semi-stable compression.
citing papers explorer
-
Realizable Bayes-Consistency for General Metric Losses
Necessary and sufficient conditions for realizable Bayes-consistency under general metric losses are the absence of infinite non-decreasing (γ_k)-Littlestone trees with γ_k → ∞.
-
Metric-valued regression
First strongly Bayes-consistent algorithm for metric-valued regression with unbounded loss in the agnostic setting, based on metric medoids and semi-stable compression.