Nearest-neighbour matching achieves usual convergence rates under general transferability conditions on source-target distribution pairs, relaxing compact support and bounded density assumptions.
Journal of Machine Learning Research , volume=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.ST 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Nearest-Neighbour Matching on Unbounded Supports and Covariate Shift Transfer
Nearest-neighbour matching achieves usual convergence rates under general transferability conditions on source-target distribution pairs, relaxing compact support and bounded density assumptions.