作者: Dmitri G. Luchinsky , Vadim N. Smelyanskiy , Andrea Duggento , Peter V. E. McClintock
DOI: 10.1103/PHYSREVE.77.061105
关键词:
摘要: A general Bayesian framework is introduced for the inference of time-varying parameters in nonstationary, nonlinear, stochastic dynamical systems. Its convergence discussed. The performance method analyzed context detecting signaling a system neurons modeled as FitzHugh-Nagumo FHN oscillators. It assumed that only fast action potentials each oscillator mixed by an unknown measurement matrix can be detected. shown proposed approach able to reconstruct unmeasured hidden variables oscillators, determine model parameters, detect stepwise changes control oscillator, and follow continuous evolution adiabatic limit.