Inferring nonlinear fractional diffusion processes from single trajectories

Abstract

We present a method to infer the arbitrary space-dependent drift and diffusion of a nonlinear stochastic model driven by multiplicative fractional Gaussian noise from a single trajectory. Our method, fractional Onsager-Machlup optimisation (fOMo), introduces a maximum likelihood estimator by minimising a field-theoretic action which we construct from the observed time series. We successfully test fOMo for a wide range of Hurst exponents using artificial data with strong nonlinearities, and apply it to a data set of daily mean temperatures. We further highlight the significant systematic estimation errors when ignoring non-Markovianity, underlining the need for nonlinear fractional inference methods when studying real-world long-range (anti-)correlated systems.

arXiv comment21 pages, 5 figures, appendices

References in corpus (24)

• Machine learning and the physical sciences
• Anomalous transport in the crowded world of biological cells
• Analysis of data sets of stochastic systems
• Power-law persistence and trends in the atmosphere: A detailed study of long temperature records
• Measurement of Anomalous Diffusion Using Recurrent Neural Networks
• On the proper reconstruction of complex dynamical systems spoilt by strong measurement noise
• Inferring the dynamics of underdamped stochastic systems
• Learning force fields from stochastic trajectories
• Rhythmicity, Recurrence, and Recovery of Flagellar Beating
• Rules of calculus in the path integral representation of white noise Langevin equations: the Onsager-Machlup approach
• High-Performance Reconstruction of Microscopic Force Fields from Brownian Trajectories
• Theoretical foundation of detrending methods for fluctuation analysis such as detrended fluctuation analysis and detrending moving average
• Stochastic actions for diffusive dynamics: Reweighting, sampling, and minimization
• A Priori Estimation Of Memory Effects In Coarse-Grained Nonlinear Systems Using The Mori-Zwanzig Formalism
• Analyzing a stochastic process driven by Ornstein-Uhlenbeck noise
• Work distribution in time-dependent logarithmic-harmonic potential: exact results and asymptotic analysis
• Estimation of drift and diffusion functions from time series data: A maximum likelihood framework
• Building general Langevin models from discrete data sets
• Path probability ratios for Langevin dynamics -- exact and approximate
• Experimental measurement of relative path probabilities and stochastic actions
• Non-parametric estimation of a Langevin model driven by correlated noise
• Stochastic action for tubes: Connecting path probabilities to measurement
• Statistical Inference of 1D Persistent Nonlinear Time Series and Application to Predictions
• Towards automated extraction and characterization of scaling regions in dynamical systems