Feedback identification of conductance-based models

This paper applies the classical prediction error method (PEM) to estimation of nonlinear discrete-time models neuronal systems subject input-additive noise. While system exhibits excitability, bifurcations, and limit-cycle oscillations, we prove consistency parameter procedure under output feedback. Hence, this provides a rigorous framework for application conventional identification methods stochastic systems. The main result exploits elementary property that conductance-based neurons have an exponentially contracting inverse dynamics. is implied by voltage-clamp experiment, which has been fundamental modeling experiment ever since pioneering work Hodgkin Huxley.