By means of Monte Carlo methods, we perform a full error analysis on the Duflo-Zucker mass model. In particular, we study the presence of correlations in the residuals to obtain a more realistic estimate of the error bars on the predicted binding energies. To further reduce the discrepancies between model prediction and experimental data we also apply a Multilayer Perceptron Neural Network. We show that the root mean square of the model further reduces of roughly 40%. We then use the resulting models to predict the composition of the outer crust of a non accreting neutron star. We provide a first estimate of the impact of error propagation on the resulting equation of state of the system.

By means of bootstrap technique, we perform a full error analysis on the Duflo-Zucker mass model. We illustrate the impact of such study on the predicted chemical composition of the outer crust of a non-accreting neutron star. We define an existence probability for each nuclear species as a function of the depth of the crust. We observe that, due to statistical uncertainties, instead of having a well defined transition between two successive layers, we have a mixture of two species.