Neural networks cannot approximate matrix inversion over domains approaching singular matrices; in safe regions they merely learn the local linear Taylor approximation of the inverse.
Dongarra, Iain S
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Rethinking Neural-based Matrix Inversion: Why can't, and Where can
Neural networks cannot approximate matrix inversion over domains approaching singular matrices; in safe regions they merely learn the local linear Taylor approximation of the inverse.