Advanced Statistical Methods to Fit Nuclear Models

Shelley, Matthew; Becker, Pierre; Gration, Amery; Pastore, Alessandro

doi:10.5506/aphyspolbsupp.12.649

Cited by 10 publications

(12 citation statements)

References 19 publications

(25 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[48]), we expect that the difference in energy per particle to drop to ≈ 20−30 keV per particle. By combining the low computational cost of ETF calculations with modern statistical technique of Gaussian Process Emulation (GPE) [51], we plan to perform a full analysis of the equation of state of the NS inner crust, including temperature effects.…”

Section: Discussionmentioning

confidence: 99%

How accurately can the Extended Thomas-Fermi method describe the inner crust of a neutron star?

Shelley,

Pastore

2020

Preprint

View full text Add to dashboard Cite

We perform a systematic comparison between the results obtained by solving fully self-consistently the Hartree-Fock-Bogoliubov equations, and those obtained using the semi-classical Extended Thomas-Fermi method, for various Wigner-Seitz cells within the inner crust of a neutron star. The lack of pairing correlations in the semi-classical approach leads to a large discrepancy between the two approaches. This discrepancy is well beyond the error of the quantum-mechanical calculation, and is related to spurious shell effects in the neutron gas.

show abstract

Section: Discussionmentioning

confidence: 99%

How accurately can the Extended Thomas-Fermi method describe the inner crust of a neutron star?

Shelley,

Pastore

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Within scientific literature there is no consensus on how to estimate error bars for NN. The standard approach based on the covariance matrix [11] can not be applied due to the typical large number of parameters and the clear difficulties in performing numerical derivatives in parameter space [12,26]. In the following, we investigate three possible methods using the example illustrated in Sec.II A.…”

Section: B Error Barsmentioning

confidence: 99%

Extrapolating from neural network models: a cautionary tale

Pastore,

Carnini

2020

Preprint

Self Cite

View full text Add to dashboard Cite

We present three different methods to estimate error bars on the predictions made using a neural network. All of them represent lower bounds for the extrapolation errors. For example, we did not include an analysis on robustness against small perturbations of the input data.At first, we illustrate the methods through a simple toy model, then, we apply them to some realistic cases related to nuclear masses. By using theoretical data simulated either with a liquiddrop model or a Skyrme energy density functional, we benchmark the extrapolation performance of the neural network in regions of the Segrè chart far away from the ones used for the training and validation. Finally, we discuss how error bars can help identifying when the extrapolation becomes too uncertain and thus unreliable.

show abstract

“…A possible alternative to NNs has been discussed in Ref. [19], and it is based on Gaussian processes (GPs) [24][25][26]. This GP method assumes that the residuals originate from some multivariate Gaussian distribution, whose covariance matrix contains some parameters to be adjusted in order to maximise the likelihood for the GP's fit to the residuals.…”

Section: Introductionmentioning

confidence: 99%

A New Mass Model for Nuclear Astrophysics: Crossing 200 keV Accuracy

Shelley

Pastore

2021

Universe

View full text Add to dashboard Cite

By using a machine learning algorithm, we present an improved nuclear mass table with a root mean square deviation of less than 200 keV. The model is equipped with statistical error bars in order to compare with available experimental data. We use the resulting model to predict the composition of the outer crust of a neutron star. By means of simple Monte Carlo methods, we propagate the statistical uncertainties of the mass model to the equation of state of the system.

show abstract

Advanced Statistical Methods to Fit Nuclear Models

Cited by 10 publications

References 19 publications

How accurately can the Extended Thomas-Fermi method describe the inner crust of a neutron star?

How accurately can the Extended Thomas-Fermi method describe the inner crust of a neutron star?

Extrapolating from neural network models: a cautionary tale

A New Mass Model for Nuclear Astrophysics: Crossing 200 keV Accuracy

Contact Info

Product

Resources

About