Some extensions and applications of self-exciting processes

Self-exciting processes allow modeling the occurrence of events whose rate depends on the history of the process. In particular, the triggering function is the term that determines the contribution of each event to the conditional intensity of the process. In this talk, we show some applications of self-exciting processes where the triggering function has a main role. First, a self-exciting point process model with change points in the triggering function is used for analyzing an earthquake dataset. Second, a self-exciting Poisson model whose outcomes are used for clustering NBA players and teams in terms of their scoring dynamics.