Randomized neural networks solve linear integro-differential equations like neutron transport by randomly fixing hidden-layer parameters and solving a convex least-squares problem for output weights, yielding competitive accuracy at lower training cost.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.NA 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Randomized Neural Networks for Integro-Differential Equations with Application to Neutron Transport
Randomized neural networks solve linear integro-differential equations like neutron transport by randomly fixing hidden-layer parameters and solving a convex least-squares problem for output weights, yielding competitive accuracy at lower training cost.