Change search
Refine search result
1 - 20 of 20
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Rows per page
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sort
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
Select
The maximal number of hits you can export is 250. When you want to export more records please use the Create feeds function.
  • 1. Capote, R.
    et al.
    Badikov, S.
    Carlson, A.
    Duran, I.
    Gunsing, F.
    Neudecker, D.
    Pronyaev, V. G.
    Schillebeeckx, P.
    Schnabel, Georg
    Uppsala University.
    Smith, D. L.
    Wallner, A.
    Unrecognized Sources of Uncertainties (USU) in Experimental Nuclear Data2019In: Nuclear Data Sheets, ISSN 0090-3752, E-ISSN 1095-9904Article in journal (Other academic)
  • 2.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    A computational EXFOR databaseIn: Article in journal (Refereed)
  • 3.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Bayesian evaluation method(s) with a treatment of unknown uncertainties2018Conference paper (Other academic)
  • 4.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Bayesian statistics with a stochastic linear model and its surprising result2018Conference paper (Other academic)
    Abstract [en]

    Bayesian methods enjoy increasing popularity in the nuclear physics community to estimate model parameters, associated uncertainties, and model bias.The occurring likelihoods are usually of simple Gaussian shape, but are intractable due to the need to run a computationally expensive code to map model parameters to predictions.For this reason, the relations between model parameters and corresponding predictions are often linearized to facilitate the analysis. Many nuclear model codes are stochastic because they simulate stochastic physical processes, such as cascades of particle collisions. Obviously, the induced stochasticity in the linear approximation needs to be taken into account in the Bayesian modeling. Surprisingly, the straight-forward Bayesian modeling of the situation can lead to strongly biased posterior estimates of the parameters with associated uncertainties that are incompatible with the true values. In this contribution, I describe the Bayesian modeling of this situation and present the results in a toy scenario where all distribution assumptions are correct. I also present my unsuccessful attempts to resolve the issue.The main aim of this contribution is to expose the issue to the Bayesian community to increase the chance of its resolution, which could have a significant impact in the field of nuclear physics.

  • 5.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Derivation of the Monte Carlo version of the marginal likelihoodManuscript (preprint) (Other academic)
  • 6.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Learning from Google: About A Computational EXFOR Database for Efficient Data Retrieval and Analysis2019Conference paper (Other academic)
    Abstract [en]

    High-level languages, such as Python and R, find broad adoption for data science and machine learning due to their expressive power and the many community-contributed packages to apply sophisticated algorithms in just a few lines of code. Despite the fast progress in these fields in recent years, the field of nuclear data evaluation remained relatively unaffected by these developments. An essential reason for this observation may be the fact that the original EXFOR format is cumbersome to deal with in highlevel languages. In this contribution, I present details about the successful conversion of the complete original EXFOR database to a NoSQL database as, e.g., employed by Google, discuss the advantages of this database architecture for nuclear data evaluation, and provide examples demonstrating the ease and flexibility of data retrieval. Finally, I show some possibilities of quick data visualization and manipulation, such as the inversion of huge experimental covariance matrices (e.g., 105×105 including correlations between data sets), underpinning the benefits of performing nuclear data evaluation in a high-level language. Conversion codes and program packages will be made available for everyone. The availability of these codes will also enable outsiders of the nuclear data field, e.g., mathematicians, statisticians, and data scientists, to test their ideas and contribute to the field.

  • 7.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Mixture GLS with examples2019Conference paper (Other academic)
  • 8.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Prototype of a pipeline for reproducible nuclear data evaluation2019Report (Other academic)
    Abstract [en]

    This document provides instructions to set up a pipeline for nuclear data evaluation using the models code TALYS. Relying on the Docker technologyfor virtualization, the pipeline is easy to install under Linux, Windows,and MacOS. In the default conguration, an evaluation of neutron-inducedcross sections of Fe-56 is performed. The pipeline can be explored and modied either using a terminal and the typical Linux command line tools ora graphical user interface accessible via a web browser. Adjustments of thepipeline enable the automatic and reproducible evaluation of other isotopes.The evaluation features automatic uncertainty correction of experimentaldata and tuning the model parameters using the Levenberg-Marquardt algorithm.The mathematical details of the evaluation are outside the scopeof this document whose focus is on the technical aspects to get the pipelinerunning. The proper working of the pipeline can be tested on a desktopcomputer. For a full scale evaluation, a multicore machine or cluster withSSH access is strongly recommended. In this mode of operation and for thetime being, Linux must be installed on the multicore machine or cluster.

  • 9.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Yet another computational EXFOR database: User manual2019Report (Other academic)
    Abstract [en]

    The EXFOR library compiled and maintained by the International Net-work of Nuclear Reaction Data Centers (NRDC) is an essential source ofinformation in the domain of nuclear physics. This document describes thesetup of the EXFOR library in the form of a MongoDB database on the localcomputer. Using the Docker technology for virtualization, the setup is easyto perform on Linux, Windows, and MacOS. Some examples to interact withthe database in the programming languages R and Python are provided. Dueto the popularity of MongoDB, equivalent functionality is also available inmany other programming languages, such as C, C++, Java, and Perl.

  • 10.
    Schnabel, Georg
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Sjöstrand, Henrik
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    A first sketch: Construction of model defect priors inspired by dynamic time warping2019In: EPJ Web of Conferences, ISSN 2101-6275, E-ISSN 2100-014X, Vol. 211, article id 07005Article in journal (Refereed)
    Abstract [en]

    Model defects are known to cause biased nuclear data evaluations if they are not taken into account in the evaluation procedure. We suggest a method to construct prior distributions for model defects for reaction models using neighboring isotopes of 56Fe as an example. A model defect is usually a function of energy and describes the difference between the model prediction and the truth. Of course, neither the truth nor the model defect are accessible. A Gaussian process (GP) enables to define a probability distribution on possible shapes of a model defect by referring to intuitively understandable concepts such as smoothness and the expected magnitude of the defect. Standard specifications of GPs impose a typical length-scale and amplitude valid for the whole energy range, which is often not justified, e.g., when the model covers both the resonance and statistical range. In this contribution, we show how a GP with energy-dependent length-scales and amplitudes can be constructed from available experimental data. The proposed construction is inspired by a technique called dynamic time warping used, e.g., for speech recognition. We demonstrate the feasibility of the data-driven determination of model defects by inferring a model defect of the nuclear models code TALYS for (n,p) reactions of isotopes with charge number between 20 and 30. The newly introduced GP parametrization besides its potential to improve evaluations for reactor relevant isotopes, such as 56Fe, may also help to better understand the performance of nuclear models in the future.

  • 11.
    Schnabel, Georg
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Sjöstrand, Henrik
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Recent ND developments and plans at UU2018In: Recent ND developments and plans at UU, 2018Conference paper (Other academic)
  • 12.
    Schnabel, Georg
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Sjöstrand, Henrik
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    VR contribution to PPPT nuclear data development: Evaluation of neutron cross section data in the fast energy range: ENR-PRD.MAT.IREMEV.7.1-T0032019Report (Other academic)
    Abstract [en]

    Fe56 is an important component in most stainless steels for fusion devices and good knowledge of Fe56 is hence paramount to forecast radiation damage effects.  An exemplary evaluation of neutron-induced reactions of Fe56 in the fast energy region has been performed and an ENDF file created. Experimental data including statistical and systematic uncertainties were extracted from the EXFOR database. Maximum likelihood optimization has been used to automatically correct systematic uncertainties of experimental data. The prior knowledge about admissive variations from the reference specification of energy-dependent TALYS parameters has been modeled as Gaussian processes whose hyperparameters have been adjusted using maximum likelihood optimization. TALYS parameters, both energy dependent and independent, were then adjusted using the Levenberg-Marquardt (LM) algorithm. The diagonal of the inverse Hessian at the posterior mode has been used as parameter covariance matrix. A sample from the multivariate normal parameter posterior distribution was given as input to a modified version of TASMAN. The TASMAN code has then be used in combination with the TEFAL code to produce the final ENDF file including covariance matrices. The work in this task proves the feasibility of the evaluation approach.

  • 13.
    Schnabel, Georg
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Sjöstrand, Henrik
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Rochman, Dimitri
    Laboratory for Reactor Physics Systems Behaviour, Paul Scherrer Institut, Villigen, Switzerland.
    J. Koning, Arjan
    IAEA.
    Interfacing TALYS with a Bayesian treatment of model defects and inconsistent data2019Conference paper (Other academic)
  • 14.
    Sjöstrand, Henrik
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Inconsistent data and uncertainties – SG44/462018Conference paper (Other academic)
  • 15.
    Sjöstrand, Henrik
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Model defect treatment for 56Fe2019Conference paper (Other academic)
  • 16.
    Sjöstrand, Henrik
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Monte Carlo integral adjustment of nuclear data libraries: experimental covariances and inconsistent data2019In: EPJ Web of Conferences, ISSN 2101-6275, E-ISSN 2100-014X, Vol. 211, no 07007Article in journal (Refereed)
    Abstract [en]

    Integral experiments can be used to adjust nuclear data libraries. Here a Bayesian Monte Carlo method based on assigning weights to the different random files is used. If the experiments are inconsistent within them-self or with the nuclear data it is shown that the adjustment procedure can lead to undesirable results. Therefore, a technique to treat inconsistent data is presented. The technique is based on the optimization of the marginal likelihood which is approximated by a sample of model calculations. The sources to the inconsistencies are discussed and the importance to take into account correlation between the different experiments is emphasized. It is found that the technique can address inconsistencies in a desirable way.

  • 17.
    Sjöstrand, Henrik
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    The identification and treatment of unrecognized uncertainties and the impact on evaluated uncertainties – SG442019Conference paper (Other academic)
  • 18.
    Sjöstrand, Henrik
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Rochman, Dimitri
    Laboratory for Reactor Physics Systems Behaviour, Paul Scherrer Institut, Villigen, Switzerland.
    Integral adjustment of nuclear data libraries: finding unrecognized systematic uncertainties and correlations2019In: Conference program & Abstract book: 2019 International Conference on Nuclear Data for Science and Technology, May 19-24, 2019, Beijing, China, 2019, p. 212-212Conference paper (Other academic)
    Abstract [sv]

    To reduce the uncertainties and obtain a better predictive power, integral adjustment of nuclear data libraries is one powerful option. Databases with integral experiments, such as the ICSBEP contain a large amount of data. When adjusting nuclear data using these integral experiments, it is important to not only include reported experimental uncertainties but also to account for the possibility of unreported experimental uncertainties and correlations between experiments, and calculation uncertainties. Unreported uncertainties and correlations can be identified and possibly quantified using marginal likelihood optimization (MLO). MLO has previously been tested for integral adjustment. In this paper, a method for including more information from the full likelihood space is pursued. It is shown that MLO can be an effective tool in addressing unknown uncertainties and correlations for a selected number of integral experiments. Results in terms of obtained parameter estimates as well as of posterior uncertainties and correlations are reported. The results are validated against an independent set of integral experiments. The findings are important for large-scale ND evaluations that heavily rely on automatization, such as TENDL, but also for any integral adjustment where a complete knowledge of all uncertainty components is out of reach. The authors believe that this is always the case.

  • 19.
    Sjöstrand, Henrik
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Rochman, Dimitri
    Laboratory for Reactor Physics Systems Behaviour, Paul Scherrer Institut, Villigen, Switzerland.
    Siefman, Daniel
    Paul Scherrer Institute (PSI), Switzerland.
    Treating inconsistent data in integral adjustment using Marginal Likelihood Optimization2019Conference paper (Other academic)
  • 20.
    Sjöstrand, Henrik
    et al.
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Schnabel, Georg
    Uppsala University, Disciplinary Domain of Science and Technology, Physics, Department of Physics and Astronomy, Applied Nuclear Physics.
    Siefman, Daniel
    EPFL/PSI.
    Treating inconsistent data in Monte Carlo integral adjustment using Marginal Likelihood Optimization2018Conference paper (Other academic)
1 - 20 of 20
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf