Online learning for multivariate hawkes processes

Yingxiang Yang, Jalal Etesami, Niao He, Negar Kiyavash

Research output: Contribution to journalConference article

Abstract

We develop a nonparametric and online learning algorithm that estimates the triggering functions of a multivariate Hawkes process (MHP). The approach we take approximates the triggering function fi,j(t) by functions in a reproducing kernel Hilbert space (RKHS), and maximizes a time-discretized version of the log-likelihood, with Tikhonov regularization. Theoretically, our algorithm achieves an O(log T) regret bound. Numerical results show that our algorithm offers a competing performance to that of the nonparametric batch learning algorithm, with a run time comparable to parametric online learning algorithms.

Original languageEnglish (US)
Pages (from-to)4938-4947
Number of pages10
JournalAdvances in Neural Information Processing Systems
Volume2017-December
StatePublished - Jan 1 2017
Event31st Annual Conference on Neural Information Processing Systems, NIPS 2017 - Long Beach, United States
Duration: Dec 4 2017Dec 9 2017

Fingerprint

Learning algorithms
Hilbert spaces

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems
  • Signal Processing

Cite this

Yang, Y., Etesami, J., He, N., & Kiyavash, N. (2017). Online learning for multivariate hawkes processes. Advances in Neural Information Processing Systems, 2017-December, 4938-4947.

Online learning for multivariate hawkes processes. / Yang, Yingxiang; Etesami, Jalal; He, Niao; Kiyavash, Negar.

In: Advances in Neural Information Processing Systems, Vol. 2017-December, 01.01.2017, p. 4938-4947.

Research output: Contribution to journalConference article

Yang, Y, Etesami, J, He, N & Kiyavash, N 2017, 'Online learning for multivariate hawkes processes', Advances in Neural Information Processing Systems, vol. 2017-December, pp. 4938-4947.
Yang Y, Etesami J, He N, Kiyavash N. Online learning for multivariate hawkes processes. Advances in Neural Information Processing Systems. 2017 Jan 1;2017-December:4938-4947.
Yang, Yingxiang ; Etesami, Jalal ; He, Niao ; Kiyavash, Negar. / Online learning for multivariate hawkes processes. In: Advances in Neural Information Processing Systems. 2017 ; Vol. 2017-December. pp. 4938-4947.
@article{e410c6a680cf44729c1b778fb6bf9ca8,
title = "Online learning for multivariate hawkes processes",
abstract = "We develop a nonparametric and online learning algorithm that estimates the triggering functions of a multivariate Hawkes process (MHP). The approach we take approximates the triggering function fi,j(t) by functions in a reproducing kernel Hilbert space (RKHS), and maximizes a time-discretized version of the log-likelihood, with Tikhonov regularization. Theoretically, our algorithm achieves an O(log T) regret bound. Numerical results show that our algorithm offers a competing performance to that of the nonparametric batch learning algorithm, with a run time comparable to parametric online learning algorithms.",
author = "Yingxiang Yang and Jalal Etesami and Niao He and Negar Kiyavash",
year = "2017",
month = "1",
day = "1",
language = "English (US)",
volume = "2017-December",
pages = "4938--4947",
journal = "Advances in Neural Information Processing Systems",
issn = "1049-5258",

}

TY - JOUR

T1 - Online learning for multivariate hawkes processes

AU - Yang, Yingxiang

AU - Etesami, Jalal

AU - He, Niao

AU - Kiyavash, Negar

PY - 2017/1/1

Y1 - 2017/1/1

N2 - We develop a nonparametric and online learning algorithm that estimates the triggering functions of a multivariate Hawkes process (MHP). The approach we take approximates the triggering function fi,j(t) by functions in a reproducing kernel Hilbert space (RKHS), and maximizes a time-discretized version of the log-likelihood, with Tikhonov regularization. Theoretically, our algorithm achieves an O(log T) regret bound. Numerical results show that our algorithm offers a competing performance to that of the nonparametric batch learning algorithm, with a run time comparable to parametric online learning algorithms.

AB - We develop a nonparametric and online learning algorithm that estimates the triggering functions of a multivariate Hawkes process (MHP). The approach we take approximates the triggering function fi,j(t) by functions in a reproducing kernel Hilbert space (RKHS), and maximizes a time-discretized version of the log-likelihood, with Tikhonov regularization. Theoretically, our algorithm achieves an O(log T) regret bound. Numerical results show that our algorithm offers a competing performance to that of the nonparametric batch learning algorithm, with a run time comparable to parametric online learning algorithms.

UR - http://www.scopus.com/inward/record.url?scp=85047008879&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85047008879&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:85047008879

VL - 2017-December

SP - 4938

EP - 4947

JO - Advances in Neural Information Processing Systems

JF - Advances in Neural Information Processing Systems

SN - 1049-5258

ER -