Constructs a computable approximating prediction region containing the full-conformal one for multi-task kernel regression in vector-valued RKHS, with theoretical volume bound for known covariance and empirical improvement over split-conformal.
Algorithmic Learning in a Random World
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
-
Approximate full-conformal multi-task regression with reproducing kernels
Constructs a computable approximating prediction region containing the full-conformal one for multi-task kernel regression in vector-valued RKHS, with theoretical volume bound for known covariance and empirical improvement over split-conformal.