Learning stochastic eigenvalues

Abstract : We train an artificial neural network with one hidden layer on realizations of the first few eigenvalues of a partial differential operator that is parameterized by a vector of independent random variables. The eigenvalues exhibit "crossings" in the high-dimensional parameter space. The training set is constructed by sampling the parameter either at random nodes or at the Smolyak collocation nodes. The performance of the neural network is evaluated empirically on a large random test set. We find that training on random or quasi-random nodes is preferable to the Smolyak nodes. The neural network outperforms the Smolyak interpolation in terms of error bias and variance on nonsimple eigenvalues but not on the simple ones.
Complete list of metadatas

Cited literature [2 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01313404
Contributor : Roman Andreev <>
Submitted on : Monday, December 12, 2016 - 5:40:19 PM
Last modification on : Tuesday, May 14, 2019 - 10:21:48 AM
Long-term archiving on : Tuesday, March 28, 2017 - 12:38:57 AM

File

UQNN_HAL_20161212.pdf
Files produced by the author(s)

Licence


Distributed under a Creative Commons Attribution 4.0 International License

Identifiers

  • HAL Id : hal-01313404, version 2

Citation

Roman Andreev. Learning stochastic eigenvalues. 2016. ⟨hal-01313404v2⟩

Share

Metrics

Record views

1414

Files downloads

732