Perturbative corrections to kernel ridge methods for NPIV that introduce eigenmode mixing to address ill-defined cases, with reported error reductions up to 99% when dimensionality parameter beta exceeds 0.7.
org/CorpusID:207063850
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Perturbative methods for non-parametric instrumental variable
Perturbative corrections to kernel ridge methods for NPIV that introduce eigenmode mixing to address ill-defined cases, with reported error reductions up to 99% when dimensionality parameter beta exceeds 0.7.