Cumulants of Hawkes Processes are Robust to Observation Noise

Multivariate Hawkes processes (MHPs) are widely used in a variety of fields to model the occurrence of causally related discrete events in continuous time. Most state-of-the-art approaches address the problem of learning MHPs from perfect traces without noise. In practice, the process through which events are collected might introduce noise in the timestamps. In this work, we address the problem of learning the causal structure of MHPs when the observed timestamps of events are subject to random and unknown shifts, also known as random translations. We prove that the cumulants of MHPs are invariant to random translations, and therefore can be used to learn their underlying causal structure. Furthermore, we empirically characterize the effect of random translations on state-of-the-art learning methods. We show that maximum likelihood-based estimators are brittle, while cumulant-based estimators remain stable even in the presence of significant time shifts.

[1]  Flavio Figueiredo,et al.  Fast Estimation of Causal Interactions using Wold Processes , 2018, NeurIPS.

[2]  Uri T Eden,et al.  A point process framework for relating neural spiking activity to spiking history, neural ensemble, and extrinsic covariate effects. , 2005, Journal of neurophysiology.

[3]  Sheldon M. Ross,et al.  Stochastic Processes , 2018, Gauge Integral Structures for Stochastic Calculus and Quantum Electrodynamics.

[4]  Bernhard Schölkopf,et al.  Uncovering the Temporal Dynamics of Diffusion Networks , 2011, ICML.

[5]  Aymen Jedidi,et al.  Long-Time Behavior of a Hawkes Process-Based Limit Order Book , 2015, SIAM J. Financial Math..

[6]  Vincent Rivoirard,et al.  Inference of functional connectivity in Neurosciences via Hawkes processes , 2013, 2013 IEEE Global Conference on Signal and Information Processing.

[7]  E. Bacry,et al.  Hawkes Processes in Finance , 2015, 1502.04592.

[8]  A. Hawkes.  Point Spectra of Some Mutually Exciting Point Processes , 1971 .

[9]  Riadh Zaatour,et al.  Hawkes Process: Fast Calibration, Application to Trade Clustering and Diffusive Limit , 2013, Journal of Futures Markets.

[10]  E. Bacry,et al.  Non-parametric kernel estimation for symmetric Hawkes processes. Application to high frequency financial data , 2011, 1112.1838.

[11]  Stefan Rotter,et al.  Cumulants of Hawkes point processes. , 2014, Physical review. E, Statistical, nonlinear, and soft matter physics.

[12]  R. Cowan.  An introduction to the theory of point processes , 1978 .

[13]  Ulrike Goldschmidt,et al.  An Introduction To The Theory Of Point Processes , 2016 .

[14]  Emmanuel Bacry,et al.  Second order statistics characterization of Hawkes processes and non-parametric estimation , 2014, 1401.0903.

[15]  Hongyuan Zha,et al.  Learning Granger Causality for Hawkes Processes , 2016, ICML.

[16]  Raja Jurdak,et al.  Modeling stochastic processes in disease spread across a heterogeneous social system , 2018, Proceedings of the National Academy of Sciences.

[17]  E. Bacry,et al.  Estimation of slowly decreasing Hawkes kernels: application to high-frequency order book dynamics , 2016 .

[18]  Felipe Gerhard,et al.  On the stability and dynamics of stochastic spiking neuron models: Nonlinear Hawkes process and point process GLMs , 2017, PLoS Comput. Biol..

[19]  Alessandro Vespignani,et al.  Predictability and epidemic pathways in global outbreaks of infectious diseases: the SARS case study , 2007, BMC medicine.

[20]  Niao He,et al.  Online Learning for Multivariate Hawkes Processes , 2017, NIPS.

[21]  Joël Chadoeuf,et al.  Estimating second order characteristics of point processes with known independent noise , 2012, Statistics and Computing.

[22]  Hongyuan Zha,et al.  Learning Hawkes Processes from Short Doubly-Censored Event Sequences , 2017, ICML.

[23]  Kun Zhang,et al.  Learning Network of Multivariate Hawkes Processes: A Time Series Approach , 2016, UAI.

[24]  Patrick Thiran,et al.  Learning Hawkes Processes Under Synchronization Noise , 2019, ICML.

[25]  Gordon J. Ross,et al.  ABC Learning of Hawkes Processes with Missing or Noisy Event Times , 2020 .

[26]  Matthias Grossglauser,et al.  Learning Hawkes Processes from a Handful of Events , 2019, NeurIPS.

[27]  Thomas Josef Liniger,et al.  Multivariate Hawkes processes , 2009 .

[28]  Constantine Caramanis,et al.  Learning Graphs from Noisy Epidemic Cascades , 2019, Abstracts of the 2019 SIGMETRICS/Performance Joint International Conference on Measurement and Modeling of Computer Systems.

[29]  Zhen Qin,et al.  Hawkes Process Inference With Missing Data , 2018, AAAI.

[30]  Emmanuel Bacry,et al.  Uncovering Causality from Multivariate Hawkes Integrated Cumulants , 2016, ICML.

[31]  L. Cucala.  Intensity Estimation for Spatial Point Processes Observed with Noise , 2008 .

[32]  R. Dahlhaus,et al.  Graphical Modeling for Multivariate Hawkes Processes with Nonparametric Link Functions , 2016, 1605.06759.

[33]  A. Hawkes.  Spectra of some self-exciting and mutually exciting point processes , 1971 .

[34]  Le Song,et al.  Learning Triggering Kernels for Multi-dimensional Hawkes Processes , 2013, ICML.