King's College London

Research portal

Critical scaling in hidden state inference for linear Langevin dynamics

Research output: Contribution to journalArticle

Standard

Critical scaling in hidden state inference for linear Langevin dynamics. / Bravi, B; Sollich, P.

In: Journal of Statistical Mechanics: Theory and Experiment, No. 6, 22.06.2017.

Research output: Contribution to journalArticle

Harvard

Bravi, B & Sollich, P 2017, 'Critical scaling in hidden state inference for linear Langevin dynamics', Journal of Statistical Mechanics: Theory and Experiment, no. 6. https://doi.org/10.1088/1742-5468

APA

Bravi, B., & Sollich, P. (2017). Critical scaling in hidden state inference for linear Langevin dynamics. Journal of Statistical Mechanics: Theory and Experiment, (6). https://doi.org/10.1088/1742-5468

Vancouver

Bravi B, Sollich P. Critical scaling in hidden state inference for linear Langevin dynamics. Journal of Statistical Mechanics: Theory and Experiment. 2017 Jun 22;(6). https://doi.org/10.1088/1742-5468

Author

Bravi, B ; Sollich, P. / Critical scaling in hidden state inference for linear Langevin dynamics. In: Journal of Statistical Mechanics: Theory and Experiment. 2017 ; No. 6.

Bibtex Download

@article{64381c9ca12b495e8f2298c2950b582b,
title = "Critical scaling in hidden state inference for linear Langevin dynamics",
abstract = "We consider the problem of inferring the dynamics of unknown (i.e. hidden)nodes from a set of observed trajectories and study analytically the average prediction error and the typical relaxation time of correlations between errors. We focus on a stochastic linear dynamics of continuous degrees of freedom interacting via random Gaussian couplings in the infinite network size limit. The expected error on the hidden time courses can be found as the equal-time hidden-to-hidden covariance of the probability distribution conditioned on observations. In the stationary regime, we analyze the phase diagram in the space of relevant parameters, namely the ratio between the numbers of observed and hidden nodes, the degree of symmetry of the interactions and the amplitudes of the hidden-to-hidden and hidden-to-observed couplings relative to the decay constant of the internal hidden dynamics. In particular, we identify critical regions in parameter space where the relaxation time and the inference error diverge, and determine the corresponding scaling behaviour.",
author = "B Bravi and P Sollich",
year = "2017",
month = "6",
day = "22",
doi = "10.1088/1742-5468",
language = "English",
journal = "Journal of Statistical Mechanics: Theory and Experiment",
issn = "1742-5468",
publisher = "IOP Publishing Ltd.",
number = "6",

}

RIS (suitable for import to EndNote) Download

TY - JOUR

T1 - Critical scaling in hidden state inference for linear Langevin dynamics

AU - Bravi, B

AU - Sollich, P

PY - 2017/6/22

Y1 - 2017/6/22

N2 - We consider the problem of inferring the dynamics of unknown (i.e. hidden)nodes from a set of observed trajectories and study analytically the average prediction error and the typical relaxation time of correlations between errors. We focus on a stochastic linear dynamics of continuous degrees of freedom interacting via random Gaussian couplings in the infinite network size limit. The expected error on the hidden time courses can be found as the equal-time hidden-to-hidden covariance of the probability distribution conditioned on observations. In the stationary regime, we analyze the phase diagram in the space of relevant parameters, namely the ratio between the numbers of observed and hidden nodes, the degree of symmetry of the interactions and the amplitudes of the hidden-to-hidden and hidden-to-observed couplings relative to the decay constant of the internal hidden dynamics. In particular, we identify critical regions in parameter space where the relaxation time and the inference error diverge, and determine the corresponding scaling behaviour.

AB - We consider the problem of inferring the dynamics of unknown (i.e. hidden)nodes from a set of observed trajectories and study analytically the average prediction error and the typical relaxation time of correlations between errors. We focus on a stochastic linear dynamics of continuous degrees of freedom interacting via random Gaussian couplings in the infinite network size limit. The expected error on the hidden time courses can be found as the equal-time hidden-to-hidden covariance of the probability distribution conditioned on observations. In the stationary regime, we analyze the phase diagram in the space of relevant parameters, namely the ratio between the numbers of observed and hidden nodes, the degree of symmetry of the interactions and the amplitudes of the hidden-to-hidden and hidden-to-observed couplings relative to the decay constant of the internal hidden dynamics. In particular, we identify critical regions in parameter space where the relaxation time and the inference error diverge, and determine the corresponding scaling behaviour.

U2 - 10.1088/1742-5468

DO - 10.1088/1742-5468

M3 - Article

JO - Journal of Statistical Mechanics: Theory and Experiment

JF - Journal of Statistical Mechanics: Theory and Experiment

SN - 1742-5468

IS - 6

ER -

View graph of relations

© 2018 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454