Department of Statistics, UCLA

A new diagnostic method for point processes is here presented. It is based on their second-order analysis, transforming the original point process by the inverse of its conditional intensity function in order to form a generalized estimate of various second-order point process properties. The result is generalized versions of the spectral density, R/S statistic, correlation integral and K-function, which can be used to test the fit of complex point process models with arbitrary conditional intensity model, rather than a stationary Poisson model.