We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

physics.comp-ph

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Physics > Computational Physics

Title: Predicting nucleation near the spinodal in the Ising model using machine learning

Abstract: We use a Convolutional Neural Network (CNN) and two logistic regression models to predict the probability of nucleation in the two-dimensional Ising model. The three models successfully predict the probability for the Nearest Neighbor Ising model for which classical nucleation is observed. The CNN outperforms the logistic regression models near the spinodal of the Long Range Ising model, but the accuracy of its predictions decreases as the quenches approach the spinodal. Occlusion analysis suggests that this decrease is due to the vanishing difference between the density of the nucleating droplet and the background. Our results are consistent with the general conclusion that predictability decreases near a critical point.
Subjects: Computational Physics (physics.comp-ph); Machine Learning (cs.LG)
Journal reference: Phys. Rev. E 103, 033305 (2021)
DOI: 10.1103/PhysRevE.103.033305
Cite as: arXiv:2004.09575 [physics.comp-ph]
  (or arXiv:2004.09575v2 [physics.comp-ph] for this version)

Submission history

From: Shan Huang [view email]
[v1] Mon, 20 Apr 2020 19:04:38 GMT (1187kb,D)
[v2] Thu, 8 Oct 2020 16:25:15 GMT (4932kb,D)

Link back to: arXiv, form interface, contact.