EVIDENCE FOR NEGLECTED LINEARITY IN NOISY CHAOTIC MODELS.

In practice it is very difficult to distinguish between stochastic and chaotic nonlinearity. It is also difficult to identify nonlinear structures in time series when the hidden dynamics are complex. In this paper, we provide evidence that in the presence of specific high-dimensional underlying structures, White [1989] and Theiler et al. [1992] tests fail to detect nonlinearity. [ABSTRACT FROM AUTHOR]