Point processes on linear networks 

Point processes on linear networks refer to a specialised area of research that studies the statistical behaviours governing events on linear network structures. The use of point processes on linear networks is particularly relevant when addressing practical research questions, such as understanding distributional behaviours of traffic accidents and street crimes. The ability to unravel these events' intricate temporal and spatial dynamics is of great importance, with significant implications for both scientific research and policy development in these areas.


In this talk, we highlight the need to consider the underlying network when studying the distributional behaviour of an underlying stochastic process supposed to have generated observed data occurring along a network structure. Then, we turn to present different kernel- and Voronoi-based intensity estimators, which are of great use to understanding how the underlying stochastic process uses space. Applications to street crimes and traffic accidents are included.