Using Machine Learning Methods to Predict Bias in Nuclear Criticality Safety

Grechanuk, Pavel Aleksandrovi; Rising, Michael Evan; Palmer, Todd

doi:10.1080/23324309.2019.1585877

Title: Using Machine Learning Methods to Predict Bias in Nuclear Criticality Safety

Abstract

The manuscript provides a description of the application of machine-learning (ML) tools to the prediction of bias in criticality safety analysis. In particular, a set of over 1000 experiments included in the Whisper package were fed into a variety of ML algorithms (notably Random Forest and AdaBoost) implemented in SciKit-Learn using k-eigenvalue sensitivities (with and without energy dependence) for individual nuclides, and optionally, the simulated k_eff as the training features. Ultimately, the ML model was used to predict the bias (k_sim - k_exp). The results indicate that use of energy-integrated sensitivity profiles with k_sim as training features led to the best predictions as quantified by root-mean square and mean absolute errors. In particular, the best-case estimates came from AdaBoost, with a mean absolute error of 0.00174, which is less than the mean experimental uncertainty of 0.00328 for the experiments included.

Authors:

Grechanuk, Pavel Aleksandrovi ^[1];

^[2]; Palmer, Todd ^[1]

Oregon State Univ., Corvallis, OR (United States)
Los Alamos National Lab. (LANL), Los Alamos, NM (United States)

Publication Date:: Thu Mar 28 00:00:00 EDT 2019

Research Org.:: Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)

Sponsoring Org.:: USDOE National Nuclear Security Administration (NNSA)

OSTI Identifier:: 1544713

Report Number(s):: LA-UR-18-24800
Journal ID: ISSN 2332-4309

Grant/Contract Number:: 89233218CNA000001

Resource Type:: Accepted Manuscript

Journal Name:: Journal of Computational and Theoretical Transport

Additional Journal Information:: Journal Volume: 47; Journal Issue: 4-6; Journal ID: ISSN 2332-4309

Publisher:: Taylor and Francis

Country of Publication:: United States

Language:: English

Subject:: 73 NUCLEAR PHYSICS AND RADIATION PHYSICS; machine learning; Monte Carlo; criticality

Citation Formats


                    Grechanuk, Pavel Aleksandrovi, Rising, Michael Evan, and Palmer, Todd. Using Machine Learning Methods to Predict Bias in Nuclear Criticality Safety.  United States: N. p., 2019. 
Web.  doi:10.1080/23324309.2019.1585877.

Copy to clipboard


                    Grechanuk, Pavel Aleksandrovi, Rising, Michael Evan, & Palmer, Todd. Using Machine Learning Methods to Predict Bias in Nuclear Criticality Safety.  United States.  https://doi.org/10.1080/23324309.2019.1585877

Copy to clipboard


                    Grechanuk, Pavel Aleksandrovi, Rising, Michael Evan, and Palmer, Todd. Thu .  
"Using Machine Learning Methods to Predict Bias in Nuclear Criticality Safety".  United States.  https://doi.org/10.1080/23324309.2019.1585877.  https://www.osti.gov/servlets/purl/1544713.

Copy to clipboard


                    
@article{osti_1544713,

  title        = {Using Machine Learning Methods to Predict Bias in Nuclear Criticality Safety},

  author       = {Grechanuk, Pavel Aleksandrovi and Rising, Michael Evan and Palmer, Todd},

  abstractNote = {The manuscript provides a description of the application of machine-learning (ML) tools to the prediction of bias in criticality safety analysis. In particular, a set of over 1000 experiments included in the Whisper package were fed into a variety of ML algorithms (notably Random Forest and AdaBoost) implemented in SciKit-Learn using k-eigenvalue sensitivities (with and without energy dependence) for individual nuclides, and optionally, the simulated keff as the training features. Ultimately, the ML model was used to predict the bias (ksim - kexp). The results indicate that use of energy-integrated sensitivity profiles with ksim as training features led to the best predictions as quantified by root-mean square and mean absolute errors. In particular, the best-case estimates came from AdaBoost, with a mean absolute error of 0.00174, which is less than the mean experimental uncertainty of 0.00328 for the experiments included.},

  doi          = {10.1080/23324309.2019.1585877},

  journal      = {Journal of Computational and Theoretical Transport},

  number       = 4-6,

  volume       = 47,

  place        = {United States},

  year         = {Thu Mar 28 00:00:00 EDT 2019},

  month        = {Thu Mar 28 00:00:00 EDT 2019}

}

Copy to clipboard

Journal Article:

Free Publicly Available Full Text

Accepted Manuscript (DOE)

Publisher's Version of Record

https://doi.org/10.1080/23324309.2019.1585877

Other availability

Search WorldCat to find libraries that may hold this journal

Citation Metrics:

Cited by: 12 works

Citation information provided by
Web of Science

Save / Share:

Export Metadata

Save to My Library

Works referenced in this record:

Sensitivity- and Uncertainty-Based Criticality Safety Validation Techniques
journal, March 2004

Broadhead, B. L.; Rearden, B. T.; Hopper, C. M.
Nuclear Science and Engineering, Vol. 146, Issue 3
DOI: 10.13182/NSE03-2

Whisper: Sensitivity/Uncertainty-Based Computational Methods and Software for Determining Baseline Upper Subcritical Limits
journal, September 2015

Kiedrowski, Brian C.; Brown, Forrest B.; Conlin, Jeremy L.
Nuclear Science and Engineering, Vol. 181, Issue 1
DOI: 10.13182/NSE14-99

A Bayesian CART algorithm
journal, June 1998

Dension, D.
Biometrika, Vol. 85, Issue 2
DOI: 10.1093/biomet/85.2.363

Perturbation Theory Eigenvalue Sensitivity Analysis with Monte Carlo Techniques
journal, March 2004

Rearden, B. T.
Nuclear Science and Engineering, Vol. 146, Issue 3
DOI: 10.13182/NSE03-03

Ensemble Methods in Machine Learning
book, January 2000

Dietterich, Thomas G.
Multiple Classifier Systems
DOI: 10.1007/3-540-45014-9_1

Random Forests
journal, January 2001

Breiman, Leo
Machine Learning, Vol. 45, Issue 1, p. 5-32
DOI: 10.1023/A:1010933404324

Similar Records in DOE PAGES and OSTI.GOV collections:

Using Machine Learning to Predict MCNP Bias

Technical Report Grechanuk, Pavel Aleksandrovi

For many real-world applications in radiation transport where simulations are compared to experimental measurements, like in nuclear criticality safety, the bias (simulated - experimental k_eff) in the calculation is an extremely important quantity used for code validation. The objective of this project is to accurately predict the bias of MCNP6 [1] criticality calculations using machine learning (ML) algorithms, with the intention of creating a tool that can complement the current nuclear criticality safety methods. In the latest release of MCNP6, the Whisper tool is available for criticality safety analysts and includes a large catalogue of experimental benchmarks, sensitivity profiles, andmore »« less
https://doi.org/10.2172/1416276

Full Text Available
Predicting compressive strength of consolidated molecular solids using computer vision and deep learning

Journal Article Gallagher, Brian ; Rever, Matthew ; Loveland, Donald ; ... - Materials & Design

We explore the application of computer vision and machine learning (ML) techniques to predict material properties (e.g., compressive strength) based on SEM images. We show that it's possible to train ML models to predict materials performance based on SEM images alone, demonstrating this capability on the real-world problem of predicting uniaxially compressed peak stress of consolidated molecular solids samples. Our image-based ML approach reduces mean absolute percentage error (MAPE) by an average of 24% over baselines representative of the current state-of-the-practice (i.e., domain-expert's analysis and correlation). We compared two complementary approaches to this problem: (1) a traditional ML approach, randommore »« less
Cited by 21
https://doi.org/10.1016/j.matdes.2020.108541
Molecular dipole moment learning via rotationally equivariant derivative kernels in molecular-orbital-based machine learning

Journal Article Sun, Jiace ; Cheng, Lixue ; Miller, Thomas F. - Journal of Chemical Physics

This study extends the accurate and transferable molecular-orbital-based machine learning (MOB-ML) approach to modeling the contribution of electron correlation to dipole moments at the cost of Hartree–Fock computations. A MOB pairwise decomposition of the correlation part of the dipole moment is applied, and these pair dipole moments could be further regressed as a universal function of MOs. The dipole MOB features consist of the energy MOB features and their responses to electric fields. An interpretable and rotationally equivariant derivative kernel for Gaussian process regression (GPR) is introduced to learn the dipole moment more efficiently. The proposed problem setup, feature design,more »« less
https://doi.org/10.1063/5.0101280

Full Text Available
Lithium-Ion Battery Diagnostics Using Electrochemical Impedance via Machine-Learning

Conference Gasper, Paul ; Schiek, Andrew ; Shimonishi, Yuta ; ...

Diagnosing battery states such as health, state-of-charge, or temperature is crucial for ensuring the safety and reliability of electrochemical energy storage systems. While some states, such as temperature, may be measured using cheap sensors, accurate diagnosis of battery health metrics usually requires time-consuming performance measurements, making them infeasible for use in real-world operation. These health metrics can be measured during lab-testing and then estimated on-line using predictive life models or via state observer algorithms such as Kalman filters, but these predictive methods should be supplemented by actual measurement of battery health whenever possible to ensure reliability. Rapid measurement of batterymore »« less
Full Text Available
Accurate Prediction of HSE06 Band Structures for a Diverse Set of Materials Using Δ-Learning

Journal Article Adhikari, Santosh ; Clary, Jacob ; Sundararaman, Ravishankar ; ... - Chemistry of Materials

Here we used machine learning (ML) to accurately predict eigenvalues of the hybrid HSE06 functional using eigenvalues computed by the less computationally expensive PBE functional and associated electronic features based on the k-point resolved atomic band character. The ML model was trained by using eigenvalues from only one k-point for each of the 168 compounds in the training set. The HSE06 eigenvalues across all k-points were then predicted for a separate set of 169 compounds with a mean absolute error (MAE) of 0.13 eV, representing a significant improvement over the error of PBE-computed eigenvalues relative to that of HSE06 (MAEmore »« less
https://doi.org/10.1021/acs.chemmater.3c01131

Similar Records

Title: Using Machine Learning Methods to Predict Bias in Nuclear Criticality Safety

Abstract

Citation Formats

Sensitivity- and Uncertainty-Based Criticality Safety Validation Techniques journal, March 2004

Whisper: Sensitivity/Uncertainty-Based Computational Methods and Software for Determining Baseline Upper Subcritical Limits journal, September 2015

A Bayesian CART algorithm journal, June 1998

Perturbation Theory Eigenvalue Sensitivity Analysis with Monte Carlo Techniques journal, March 2004

Ensemble Methods in Machine Learning book, January 2000

Random Forests journal, January 2001

Sensitivity- and Uncertainty-Based Criticality Safety Validation Techniques
journal, March 2004

Whisper: Sensitivity/Uncertainty-Based Computational Methods and Software for Determining Baseline Upper Subcritical Limits
journal, September 2015

A Bayesian CART algorithm
journal, June 1998

Perturbation Theory Eigenvalue Sensitivity Analysis with Monte Carlo Techniques
journal, March 2004

Ensemble Methods in Machine Learning
book, January 2000

Random Forests
journal, January 2001