PulseAugur
LIVE 06:15:48
tool · [1 source] ·
0
tool

Researchers develop parallel algorithm for faster Hawkes process inference

Researchers have developed a massively parallel algorithm for estimating multivariate Hawkes processes, a class of self-exciting point processes. Their method leverages sparse transition matrices and parallel prefix scans to achieve a computational complexity of approximately O(N/P) with P processors, significantly speeding up calculations. This approach computes the exact likelihood without approximations and has demonstrated orders-of-magnitude speedups on large datasets, with an open-source PyTorch library available. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a novel, highly parallelizable inference method for self-exciting point processes, potentially impacting time-series analysis and event prediction in AI applications.

RANK_REASON Academic paper detailing a new computational method for Hawkes processes. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

COVERAGE [1]

  1. arXiv cs.LG TIER_1 · Ahmer Raza, Hudson Smith ·

    Massively Parallel Exact Inference for Hawkes Processes

    arXiv:2604.01342v2 Announce Type: replace Abstract: Multivariate Hawkes processes are a widely used class of self-exciting point processes, but maximum likelihood estimation naively scales as $O(N^2)$ in the number of events. The canonical linear exponential Hawkes process admits…