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: Revisiting the dynamics of Bose-Einstein condensates in a double well by deep learning with a hybrid network

Abstract: Deep learning, accounting for the use of an elaborate neural network, has recently been developed as an efficient and powerful tool to solve diverse problems in physics and other sciences. In the present work, we propose a novel learning method based on a hybrid network integrating two different kinds of neural networks: Long Short-Term Memory(LSTM) and Deep Residual Network(ResNet), in order to overcome the difficulty met in numerically simulating strongly-oscillating dynamical evolutions of physical systems. By taking the dynamics of Bose-Einstein condensates in a double-well potential as an example, we show that our new method makes a high efficient pre-learning and a high-fidelity prediction about the whole dynamics. This benefits from the advantage of the combination of the LSTM and the ResNet and is impossibly achieved by a single network in the case of direct learning. Our method can be applied for simulating complex cooperative dynamics in a system with fast multiple-frequency oscillations with the aid of auxiliary spectrum analysis.
Comments: 11 pages, 8 figures
Subjects: Computational Physics (physics.comp-ph); Machine Learning (cs.LG); Quantum Physics (quant-ph)
DOI: 10.1007/s11467-021-1111-8
Cite as: arXiv:2104.14657 [physics.comp-ph]
  (or arXiv:2104.14657v2 [physics.comp-ph] for this version)

Submission history

From: Jing Qian [view email]
[v1] Sun, 25 Apr 2021 07:51:43 GMT (1291kb)
[v2] Mon, 1 Nov 2021 13:00:38 GMT (1601kb)

Link back to: arXiv, form interface, contact.