Learning Hawkes Processes from Short Doubly-Censored Event Sequences

TL;DR

This paper introduces a data synthesis approach for learning Hawkes processes from short, incomplete event sequences, significantly improving model training on such data.

Contribution

It proposes a novel sampling-stitching data synthesis method to effectively handle short doubly-censored event sequences in Hawkes process learning.

Findings

Improves learning accuracy for both time-invariant and time-varying Hawkes processes.

Demonstrates effectiveness on synthetic and real-world datasets.

Enhances model robustness with incomplete event data.

Abstract

Many real-world applications require robust algorithms to learn point processes based on a type of incomplete data --- the so-called short doubly-censored (SDC) event sequences. We study this critical problem of quantitative asynchronous event sequence analysis under the framework of Hawkes processes by leveraging the idea of data synthesis. Given SDC event sequences observed in a variety of time intervals, we propose a sampling-stitching data synthesis method --- sampling predecessors and successors for each SDC event sequence from potential candidates and stitching them together to synthesize long training sequences. The rationality and the feasibility of our method are discussed in terms of arguments based on likelihood. Experiments on both synthetic and real-world data demonstrate that the proposed data synthesis method improves learning results indeed for both time-invariant and time-varying Hawkes processes.