On the use CARMA(p,q) models for the intensity of point processes
Conference
64th ISI World Statistics Congress - Ottawa, Canada
Format: IPS Abstract
Keywords: estimation, hawkes, simulation
Session: IPS 172 - Theoretical and computational developments of modeling non-Gaussian stochastic processes
Thursday 20 July 2 p.m. - 3:40 p.m. (Canada/Eastern)
Abstract
Point processes are useful to describe the dynamics of observed event times. A standard Hawkes model with an exponential kernel implies a strictly decreasing behavior of the autocorrelation function. As we empirically reject the monotonicity assumption on the autocorrelation function, we use a CARMA(p,q) model for the intensity of the Hawkes process. Our model, that can be seen as a generalization of the Hawkes process with exponential kernel, is able to reproduce more realistic dependence structures.