Fast approximate Bayesian inference for stable differential equation models

Inference for mechanistic models is challenging because of nonlinear interactions between model parameters and a lack of identifiability. Here we focus on a specific class of mechanistic models, which we term stable differential equations. The dynamics in these models are approximately linear around a stable fixed point of the system. We exploit this property to develop fast approximate methods for posterior inference. We illustrate our approach using simulated data on a mechanistic neuroscience model with EEG data. More generally, stable differential equation models and the corresponding inference methods are useful for analysis of stationary time-series data. Compared to the existing state-of-the art, our methods are several orders of magnitude faster, and are particularly suited to analysis of long time-series (>10,000 time-points) and models of moderate dimension (10-50 state variables and 10-50 parameters.)

[1]  N. Rashevsky,et al.  Mathematical biology , 1961, Connecticut medicine.

[2]  R. FitzHugh Impulses and Physiological States in Theoretical Models of Nerve Membrane. , 1961, Biophysical journal.

[3]  A. Kelley The stable, center-stable, center, center-unstable, unstable manifolds , 1967 .

[4]  A. J. Hermans,et al.  A model of the spatial-temporal characteristics of the alpha rhythm. , 1982, Bulletin of mathematical biology.

[5]  D. B. Preston Spectral Analysis and Time Series , 1983 .

[6]  N. Gordon,et al.  Novel approach to nonlinear/non-Gaussian Bayesian state estimation , 1993 .

[7]  Adrian Raftery,et al.  The Number of Iterations, Convergence Diagnostics and Generic Metropolis Algorithms , 1995 .

[8]  David S. Stoffer,et al.  Time series analysis and its applications , 2000 .

[9]  K Lehnertz,et al.  Indications of nonlinear deterministic and finite-dimensional structures in time series of brain electrical activity: dependence on recording region and brain state. , 2001, Physical review. E, Statistical, nonlinear, and soft matter physics.

[10]  John R. Terry,et al.  Detection and description of non-linear interdependence in normal multichannel human EEG data , 2002, Clinical Neurophysiology.

[11]  Mathew P. Dafilis,et al.  A spatially continuous mean field theory of electrocortical activity , 2002, Network.

[12]  E. Marder,et al.  Failure of averaging in the construction of a conductance-based neuron model. , 2002, Journal of neurophysiology.

[13]  D. Liley,et al.  Modeling the effects of anesthesia on the electroencephalogram. , 2005, Physical review. E, Statistical, nonlinear, and soft matter physics.

[14]  Karl J. Friston,et al.  Modelling event-related responses in the brain , 2005, NeuroImage.

[15]  Eugene M. Izhikevich,et al.  Dynamical Systems in Neuroscience: The Geometry of Excitability and Bursting , 2006 .

[16]  J. NAGUMOt,et al.  An Active Pulse Transmission Line Simulating Nerve Axon , 2006 .

[17]  M. Giles An extended collection of matrix derivative results for forward and reverse mode algorithmic dieren tiation , 2008 .

[18]  Raymond J. Dolan,et al.  Dynamic causal models of steady-state responses , 2009, NeuroImage.

[19]  E. Meerbach Off- and online detection of dynamical phases in time series , 2009 .

[20]  A. Doucet,et al.  Particle Markov chain Monte Carlo methods , 2010 .

[21]  David Terman,et al.  Mathematical foundations of neuroscience , 2010 .

[22]  Radford M. Neal MCMC Using Hamiltonian Dynamics , 2011, 1206.1901.

[23]  M. Girolami,et al.  Riemann manifold Langevin and Hamiltonian Monte Carlo methods , 2011, Journal of the Royal Statistical Society: Series B (Statistical Methodology).

[24]  Karl J. Friston,et al.  Dynamic causal modeling with neural fields , 2012, NeuroImage.

[25]  Omiros Papaspiliopoulos,et al.  Markov chain Monte Carlo approach to parameter estimation in the FitzHugh-Nagumo model. , 2012, Physical review. E, Statistical, nonlinear, and soft matter physics.

[26]  Christophe Andrieu,et al.  Online Bayesian Inference in Some Time-Frequency Representations of Non-Stationary Processes , 2013, IEEE Transactions on Signal Processing.

[27]  Eberhard F. Kochs,et al.  Non-stationarity of EEG during wakefulness and anaesthesia: advantages of EEG permutation entropy monitoring , 2014, Journal of Clinical Monitoring and Computing.

[28]  Umberto Picchini Inference for SDE Models via Approximate Bayesian Computation , 2012, 1204.5459.

[29]  Arnaud Doucet,et al.  On Particle Methods for Parameter Estimation in State-Space Models , 2014, 1412.8695.

[30]  Andrew Gelman,et al.  The No-U-turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo , 2011, J. Mach. Learn. Res..

[31]  Johan Dahlin,et al.  Particle Metropolis–Hastings using gradient and Hessian information , 2013, Statistics and Computing.

[32]  Andrew Golightly,et al.  Delayed acceptance particle MCMC for exact inference in stochastic kinetic models , 2014, Stat. Comput..

[33]  William D. Penny,et al.  Annealed Importance Sampling for Neural Mass Models , 2016, PLoS Comput. Biol..

[34]  P. Robinson,et al.  Real-time automated EEG tracking of brain states using neural field theory , 2016, Journal of Neuroscience Methods.

[35]  Pierre Alquier,et al.  Noisy Monte Carlo: convergence of Markov chains with approximate transition kernels , 2014, Statistics and Computing.

[36]  M. Vihola,et al.  Importance sampling type correction of Markov chain Monte Carlo and exact approximations , 2016 .

[37]  Robert H. Shumway,et al.  Time series analysis and its applications : with R examples , 2017 .