Advanced search
1 file | 2.15 MB Add to list

ReaxFF parameter optimization with Monte-Carlo and evolutionary algorithms : guidelines and insights

Author
Organization
Project
  • DEFNET (DEFect NETwork materials science and engineering)
Abstract
ReaxFF is a computationally efficient force field to simulate complex reactive dynamics in extended molecular models with diverse chemistries, if reliable force-field parameters are available for the chemistry of interest. If not, they must be optimized by minimizing the error ReaxFF makes on a relevant training set. Because this optimization is far from trivial, many methods, in particular, genetic algorithms (GAs), have been developed to search for the global optimum in parameter space. Recently, two alternative parameter calibration techniques were proposed, that is, Monte-Carlo force field optimizer (MCFF) and covariance matrix adaptation evolutionary strategy (CMA-ES). In this work, CMA-ES, MCFF, and a GA method (OGOLEM) are systematically compared using three training sets from the literature. By repeating optimizations with different random seeds and initial parameter guesses, it is shown that a single optimization run with any of these methods should not be trusted blindly: nonreproducible, poor or premature convergence is a common deficiency. GA shows the smallest risk of getting trapped into a local minimum, whereas CMA-ES is capable of reaching the lowest errors for two-third of the cases, although not systematically. For each method, we provide reasonable default settings, and our analysis offers useful guidelines for their usage in future work. An important side effect impairing parameter optimization is numerical noise. A detailed analysis reveals that it can be reduced, for example, by using exclusively unambiguous geometry optimization in the training set. Even without this noise, many distinct near-optimal parameter vectors can be found, which opens new avenues for improving the training set and detecting overfitting artifacts.
Keywords
REACTIVE FORCE-FIELD, MOLECULAR-DYNAMICS SIMULATIONS, PARALLEL OPTIMIZATION, GLOBAL OPTIMIZATION, COMBUSTION, POTENTIALS, CHALLENGES, STRATEGIES, MECHANICS, FRAMEWORK

Downloads

  • article.pdf
    • full text (Author's original)
    • |
    • open access
    • |
    • PDF
    • |
    • 2.15 MB

Citation

Please use this url to cite or link to this publication:

MLA
Shchygol, Ganna, et al. “ReaxFF Parameter Optimization with Monte-Carlo and Evolutionary Algorithms : Guidelines and Insights.” JOURNAL OF CHEMICAL THEORY AND COMPUTATION, vol. 15, no. 12, 2019, pp. 6799–812, doi:10.1021/acs.jctc.9b00769.
APA
Shchygol, G., Yakovlev, A., Trnka, T., van Duin, A. C., & Verstraelen, T. (2019). ReaxFF parameter optimization with Monte-Carlo and evolutionary algorithms : guidelines and insights. JOURNAL OF CHEMICAL THEORY AND COMPUTATION, 15(12), 6799–6812. https://doi.org/10.1021/acs.jctc.9b00769
Chicago author-date
Shchygol, Ganna, Alexei Yakovlev, Tomáš Trnka, Adri CT van Duin, and Toon Verstraelen. 2019. “ReaxFF Parameter Optimization with Monte-Carlo and Evolutionary Algorithms : Guidelines and Insights.” JOURNAL OF CHEMICAL THEORY AND COMPUTATION 15 (12): 6799–6812. https://doi.org/10.1021/acs.jctc.9b00769.
Chicago author-date (all authors)
Shchygol, Ganna, Alexei Yakovlev, Tomáš Trnka, Adri CT van Duin, and Toon Verstraelen. 2019. “ReaxFF Parameter Optimization with Monte-Carlo and Evolutionary Algorithms : Guidelines and Insights.” JOURNAL OF CHEMICAL THEORY AND COMPUTATION 15 (12): 6799–6812. doi:10.1021/acs.jctc.9b00769.
Vancouver
1.
Shchygol G, Yakovlev A, Trnka T, van Duin AC, Verstraelen T. ReaxFF parameter optimization with Monte-Carlo and evolutionary algorithms : guidelines and insights. JOURNAL OF CHEMICAL THEORY AND COMPUTATION. 2019;15(12):6799–812.
IEEE
[1]
G. Shchygol, A. Yakovlev, T. Trnka, A. C. van Duin, and T. Verstraelen, “ReaxFF parameter optimization with Monte-Carlo and evolutionary algorithms : guidelines and insights,” JOURNAL OF CHEMICAL THEORY AND COMPUTATION, vol. 15, no. 12, pp. 6799–6812, 2019.
@article{8637024,
  abstract     = {ReaxFF is a computationally efficient force field to simulate complex reactive dynamics in extended molecular models with diverse chemistries, if reliable force-field parameters are available for the chemistry of interest. If not, they must be optimized by minimizing the error ReaxFF makes on a relevant training set. Because this optimization is far from trivial, many methods, in particular, genetic algorithms (GAs), have been developed to search for the global optimum in parameter space. Recently, two alternative parameter calibration techniques were proposed, that is, Monte-Carlo force field optimizer (MCFF) and covariance matrix adaptation evolutionary strategy (CMA-ES). In this work, CMA-ES, MCFF, and a GA method (OGOLEM) are systematically compared using three training sets from the literature. By repeating optimizations with different random seeds and initial parameter guesses, it is shown that a single optimization run with any of these methods should not be trusted blindly: nonreproducible, poor or premature convergence is a common deficiency. GA shows the smallest risk of getting trapped into a local minimum, whereas CMA-ES is capable of reaching the lowest errors for two-third of the cases, although not systematically. For each method, we provide reasonable default settings, and our analysis offers useful guidelines for their usage in future work. An important side effect impairing parameter optimization is numerical noise. A detailed analysis reveals that it can be reduced, for example, by using exclusively unambiguous geometry optimization in the training set. Even without this noise, many distinct near-optimal parameter vectors can be found, which opens new avenues for improving the training set and detecting overfitting artifacts.},
  author       = {Shchygol, Ganna and Yakovlev, Alexei and Trnka, Tomáš and van Duin, Adri CT and Verstraelen, Toon},
  issn         = {1549-9618},
  journal      = {JOURNAL OF CHEMICAL THEORY AND COMPUTATION},
  keywords     = {REACTIVE FORCE-FIELD,MOLECULAR-DYNAMICS SIMULATIONS,PARALLEL OPTIMIZATION,GLOBAL OPTIMIZATION,COMBUSTION,POTENTIALS,CHALLENGES,STRATEGIES,MECHANICS,FRAMEWORK},
  language     = {eng},
  number       = {12},
  pages        = {6799--6812},
  title        = {ReaxFF parameter optimization with Monte-Carlo and evolutionary algorithms : guidelines and insights},
  url          = {http://dx.doi.org/10.1021/acs.jctc.9b00769},
  volume       = {15},
  year         = {2019},
}

Altmetric
View in Altmetric
Web of Science
Times cited: