Optimal sampling of dynamical large deviations via matrix product states.

@article{Causer2021OptimalSO,
  title={Optimal sampling of dynamical large deviations via matrix product states.},
  author={Luke Causer and Mar{\'i} Carmen Ba{\~n}uls and Juan P. Garrahan},
  journal={Physical review. E},
  year={2021},
  volume={103 6-1},
  pages={
          062144
        }
}
The large deviation statistics of dynamical observables is encoded in the spectral properties of deformed Markov generators. Recent works have shown that tensor network methods are well suited to compute accurately the relevant leading eigenvalues and eigenvectors. However, the efficient generation of the corresponding rare trajectories is a harder task. Here, we show how to exploit the matrix product state approximation of the dominant eigenvector to implement an efficient sampling scheme… 

Figures from this paper

Markov Modulated Process to Model Human Mobility
We introduce a Markov Modulated Process (MMP) to describe human mobility. We represent the mobility process as a timevarying graph, where a link specifies a connection between two nodes (humans) at
A large deviation theory perspective on nanoscale transport phenomena
Understanding transport processes in complex nanoscale systems, like ionic conductivities in nanofluidic devices or heat conduction in low-dimensional solids, poses the problem of examining
Finite time large deviations via matrix product states
Luke Causer, 2 Mari Carmen Bañuls, 4 and Juan P. Garrahan 2 School of Physics and Astronomy, University of Nottingham, Nottingham, NG7 2RD, UK Centre for the Mathematics and Theoretical Physics of
Reinforcement learning of rare diffusive dynamics
TLDR
This paper presents a probabilistic simulation of the response of the immune system to laser-spot assisted, proton-proton collisions and shows clear patterns of response to different types of lasers.

References

SHOWING 1-10 OF 80 REFERENCES
Phys
  • Rep. 478, 1
  • 2009
Phys
  • Rev. E 102, 052132
  • 2020
Phys
  • Rev. Lett. 123, 200601
  • 2019
Physica A 418
  • 17
  • 2015
K
  • van Duijvendijk, and F. van Wijland, J. Phys. A 42, 75007
  • 2009
Adv
  • Phys. 57, 143
  • 2008
A 40
  • R333
  • 2007
Dynamical large deviations of two-dimensional kinetically constrained models using a neural-network state ansatz
TLDR
The scaled cumulant-generating function for the dynamical activity of the Fredrickson-Andersen model is obtained in one and two dimensions, and the first size-scaling analysis of the dynamicals activity in two dimensions is presented.
New J
  • Phys. 23, 013013
  • 2021
Eur
  • Phys. J. B 93, 74
  • 2020
...
1
2
3
4
5
...