Dynamic Hawkes Processes for Discovering Time-evolving Communities' States behind Diffusion Processes

被引:2
|
作者
Okawa, Maya [1 ,3 ]
Iwata, Tomoharu [2 ]
Tanaka, Yusuke [2 ]
Toda, Hiroyuki [1 ]
Kurashima, Takeshi [1 ]
Kashima, Hisashi [3 ]
机构
[1] NTT Corp, NTT Serv Evolut Labs, Tokyo, Japan
[2] NTT Corp, NTT Commun Sci Labs, Tokyo, Japan
[3] Kyoto Univ, Dept Intelligence Sci & Technol, Kyoto, Japan
来源
KDD '21: PROCEEDINGS OF THE 27TH ACM SIGKDD CONFERENCE ON KNOWLEDGE DISCOVERY & DATA MINING | 2021年
关键词
Hawkes processes; Event prediction; Neural networks; POINT; MODELS;
D O I
10.1145/3447548.3467248
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Sequences of events including infectious disease outbreaks, social network activities, and crimes are ubiquitous and the data on such events carry essential information about the underlying diffusion processes between communities (e.g., regions, online user groups). Modeling diffusion processes and predicting future events are crucial in many applications including epidemic control, viral marketing, and predictive policing. Hawkes processes offer a central tool for modeling the diffusion processes, in which the influence from the past events is described by the triggering kernel. However, the triggering kernel parameters, which govern how each community is influenced by the past events, are assumed to be static over time. In the real world, the diffusion processes depend not only on the influences from the past, but also the current (time-evolving) states of the communities, e.g., people's awareness of the disease and people's current interests. In this paper, we propose a novel Hawkes process model that is able to capture the underlying dynamics of community states behind the diffusion processes and predict the occurrences of events based on the dynamics. Specifically, we model the latent dynamic function that encodes these hidden dynamics by a mixture of neural networks. Then we design the triggering kernel using the latent dynamic function and its integral. The proposed method, termed DHP (Dynamic Hawkes Processes), offers a flexible way to learn complex representations of the time-evolving communities' states, while at the same time it allows to computing the exact likelihood, which makes parameter learning tractable. Extensive experiments on four real-world event datasets show that DHP outperforms five widely adopted methods for event prediction.
引用
收藏
页码:1276 / 1286
页数:11
相关论文
共 50 条
  • [41] DIFFUSION PROCESSES IN A SMALL TIME INTERVAL
    VARADHAN, SR
    COMMUNICATIONS ON PURE AND APPLIED MATHEMATICS, 1967, 20 (04) : 659 - &
  • [42] STOCHASTIC DYNAMIC INTERPOLATION FORMULA TO TIME-REVERSIBLE DIFFUSION-PROCESSES
    MISAWA, T
    JOURNAL OF MATHEMATICAL PHYSICS, 1993, 34 (02) : 775 - 784
  • [43] Time-dependent dynamic diffusion processes in coal: Model development and analysis
    Liu, Ting
    Lin, Baiquan
    INTERNATIONAL JOURNAL OF HEAT AND MASS TRANSFER, 2019, 134 : 1 - 9
  • [44] Pricing and hedging foreign equity options under Hawkes jump-diffusion processes
    Ma, Yong
    Pan, Dongtao
    Shrestha, Keshab
    Xu, Weidong
    PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS, 2020, 537
  • [45] METASTABLE STATES IN DIFFUSION-CONTROLLED PROCESSES
    BURLATSKII, SF
    PRONIN, KA
    JOURNAL OF PHYSICS A-MATHEMATICAL AND GENERAL, 1989, 22 (05): : 531 - 535
  • [46] Steady states of radially symmetric diffusion processes
    Li, Y
    Yin, JX
    ZEITSCHRIFT FUR ANGEWANDTE MATHEMATIK UND PHYSIK, 1999, 50 (06): : 849 - 859
  • [47] Steady states of radially symmetric diffusion processes
    Y. Li
    J. Yin
    Zeitschrift für angewandte Mathematik und Physik ZAMP, 1999, 50 : 849 - 859
  • [48] ESTIMATING COVID-19 TRANSMISSION TIME USING HAWKES POINT PROCESSES
    Schoenberg, Frederic
    ANNALS OF APPLIED STATISTICS, 2023, 17 (04): : 3349 - 3362
  • [49] Intentional Process Mining: Discovering and Modeling the Goals Behind Processes using Supervised Learning
    Deneckere, Rebecca
    Hug, Charlotte
    Khodabandelou, Ghazaleh
    Salinesi, Camille
    INTERNATIONAL JOURNAL OF INFORMATION SYSTEM MODELING AND DESIGN, 2014, 5 (04) : 22 - 47
  • [50] Dynamic recrystallization of garnet and related diffusion processes
    Bestmann, Michel
    Habler, Gerlinde
    Heidelbach, Florian
    Thoeni, Martin
    JOURNAL OF STRUCTURAL GEOLOGY, 2008, 30 (06) : 777 - 790