The neural hawkes process: a neurally self-modulating multivariate point process

Hongyuan Mei, Jason Eisner

Neural Information Processing Systems · 2017 · 213 citations · 22 references

Concepts

TL;DR

Events in the world can be stochastically excited or inhibited by patterns in the sequence of prior events. The study aims to discover such patterns to predict which event type will occur next and when. The authors introduce a neurally self‑modulating multivariate point process in which event intensities evolve via a continuous‑time LSTM that conditions future intensities on the recurrent network’s hidden state. The model shows desirable qualitative behavior and achieves competitive likelihood and predictive accuracy on real and synthetic datasets, even under missing‑data conditions.

Abstract

Many events occur in the world. Some event types are stochastically excited or inhibitedߞin the sense of having their probabilities elevated or decreasedߞby patterns in the sequence of previous events. Discovering such patterns can help us predict which type of event will happen next and when. We model streams of discrete events in continuous time, by constructing a neurally self-modulating multivariate point process in which the intensities of multiple event types evolve according to a novel continuous-time LSTM. This generative model allows past events to influence the future in complex and realistic ways, by conditioning future event intensities on the hidden state of a recurrent neural network that has consumed the stream of past events. Our model has desirable qualitative properties. It achieves competitive likelihood and predictive accuracy on real and synthetic datasets, including under missing-data conditions.

References

22