A new approach for point process diagnostics is presented. The method isbased on extending second-order statistics for point processes by weighting each pointby the inverse of the conditional intensity function at the point’s location. The resultis generalized versions of the spectral density, R/S statistic, correlation integral andK-function, which can be used to test the fit of a complex point process model withan arbitrary conditional intensity function, rather than a stationary Poisson model.Asymptotic properties of these generalized second-order statistics are derived, usingan approach based on martingale theory.
|Numero di pagine||20|
|Rivista||Annals of the Institute of Statistical Mathematics|
|Stato di pubblicazione||Published - 2009|
All Science Journal Classification (ASJC) codes
- Statistics and Probability