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.

arXiv comment11 pages, 8 figures

References in corpus (35)

• Adam: A Method for Stochastic Optimization
• Solving the Quantum Many-Body Problem with Artificial Neural Networks
• Direct Observation of Tunneling and Nonlinear Self-Trapping in a single Bosonic Josephson Junction
• Machine learning phases of matter
• Do Deep Nets Really Need to be Deep?
• Quantum Coherent Atomic Tunneling between Two Trapped Bose-Einstein Condensates
• Many-body quantum state tomography with neural networks
• Learning phase transitions by confusion
• Atom interferometry with Bose-Einstein condensates in a double-well potential
• Quantum-enhanced machine learning
• Machine Learning Topological Invariants with Neural Networks
• Recurrent Neural Network Wave Functions
• Spatial fragmentation of a Bose-Einstein condensate in a double-well potential
• Machine Learning for Precise Quantum Measurement
• Josephson effect in fermionic superfluids across the BEC-BCS crossover
• Quantum Adversarial Machine Learning
• Deep Learning for Feynman's Path Integral in Strong-Field Time-Dependent Dynamics
• Crossing Over from Attractive to Repulsive Interactions in a Tunneling Bosonic Josephson Junction
• Machine learning non-Markovian quantum dynamics
• Ordering of small particles in one-dimensional coherent structures by time-periodic flows
• Super-resolving the Ising model with convolutional neural networks
• Practical phase-modulation stabilization in quantum key distribution via machine learning
• Double-well magnetic trap for Bose-Einstein condensates
• Direct Detection of Dark Matter Substructure in Strong Lens Images with Convolutional Neural Networks
• Analytical Pendulum Model for a Bosonic Josephson Junction
• SpecTr: Spectral Transformer for Hyperspectral Pathology Image Segmentation
• Chaotic behavior in Casimir oscillators: A case study for phase change materials
• Nonequilibrium Thermodynamics of Restricted Boltzmann Machines
• Dynamics in asymmetric double-well condensates
• Non-linear mixing of Bogoliubov modes in a bosonic Josephson junction
• Deep-learning-assisted detection and termination of spiral- and broken-spiral waves in mathematical models for cardiac tissue
• Effective three-body interactions for bosons in a double-well confinement
• Effective two-mode description of a few ultra-cold bosons in a double-well potential
• Generation of Bose-Einstein Condensates' Ground State Through Machine Learning
• Quantum dynamics in potentials with fast spatial oscillations