[2410.07430] EventFlow: Forecasting Temporal Point Processes with Flow Matching

Computer Science > Machine Learning arXiv:2410.07430 (cs) [Submitted on 9 Oct 2024 (v1), last revised 6 Apr 2026 (this version, v3)] Title:EventFlow: Forecasting Temporal Point Processes with Flow Matching Authors:Gavin Kerrigan, Kai Nelson, Padhraic Smyth View a PDF of the paper titled EventFlow: Forecasting Temporal Point Processes with Flow Matching, by Gavin Kerrigan and 2 other authors View PDF HTML (experimental) Abstract:Continuous-time event sequences, in which events occur at irregular intervals, are ubiquitous across a wide range of industrial and scientific domains. The contemporary modeling paradigm is to treat such data as realizations of a temporal point process, and in machine learning it is common to model temporal point processes in an autoregressive fashion using a neural network. While autoregressive models are successful in predicting the time of a single subsequent event, their performance can degrade when forecasting longer horizons due to cascading errors and myopic predictions. We propose EventFlow, a non-autoregressive generative model for temporal point processes. The model builds on the flow matching framework in order to directly learn joint distributions over event times, side-stepping the autoregressive process. EventFlow is simple to implement and achieves a 20%-53% lower forecast error than the nearest baseline on standard TPP benchmarks while simultaneously using fewer model calls at sampling time. Comments: Subjects: Machine Learning (cs.L...