Nonlinear mapping between continuous- and discrete-time dynamics.

Linear mapping is widely used in dynamic modeling and empirical data analysis, but it suffers from the serious shortcoming that it does not work in the common case of time intervals being large. The fundamental cause of the failure is that the linear mapping does not take into account the coupling effects of multiple events within a discrete time interval. Here, we develop a theoretical framework to provide a nonlinear mapping between continuous- and discrete-time dynamics by accounting for the coupling effect. We have verified the effectiveness of our mapping by exploring classical susceptible-infected-susceptible and susceptible-infected-recovered models. In particular, we give a quantitative criterion that the sum of two transition probabilities â€"from one state to the other and vice versaâ€" must be strictly less than 1 for binary-state dynamics. [ABSTRACT FROM AUTHOR]