Center for Advanced Intelligence Project

Dr. Young Lee (Data61/Australian National University, Australia)

Title:
Inference for interval censored point processes with random intensities

Abstract:
A problem frequently encountered in point process modelling is the fact that event times are usually not known. The only available information is the number of events over a given interval.

In most applications concerning point processes, we are generally confronted with the problem that we do not have the exact times at which events have taken place. In such cases, the only information we have is the number of events over a given interval. For example, there can be many automobile accidents per day in Australia, and it is common to record the number of accidents for that day, rather than the exact times at which these events happened.

The purpose of this talk is to give an overview of some recent techniques to perform inference in these circumstances. We focus on one-dimensional point processes with random intensities.