Point Process

Point processes are mathematical models used to describe the timing and characteristics of events occurring in continuous time, with applications ranging from neuroscience to finance. Current research focuses on developing more efficient and flexible models, including those based on transformer networks and Bayesian methods, to address challenges like high-dimensionality and the computational cost of inference. These advancements are improving the accuracy and interpretability of point process models, leading to better predictions and insights in diverse fields requiring analysis of event sequences. The development of integration-free methods and improved algorithms for handling various data types (e.g., spatiotemporal, marked, interval-censored) is a significant area of progress.