Ram, Ashwin and Srijith, P K
(2017)
Accelerating Hawkes Process for Modelling Event History Data.
In: ICML 2017 Time Series Workshop, Sydney, Australia.
Abstract
Hawkes Processes are probabilistic models use-
ful for modelling the occurrences of events over
time. They exhibit mutual excitation property,
where a past event influences future events. This
has been successful in modelling the evolution
of memes and user behaviour in social net-
works. In the Hawkes process, the occurrences
of events are determined by an underlying inten-
sity function which considers the influence from
past events. The intensity function models the
mutual-exciting nature by adding up the influ-
ence from past events. The calculation of the in-
tensity function for every new event requires time
proportional to the number of past events. When
the number of events is high, the repeated in-
tensity function calculation will become expen-
sive. We develop a faster approach which takes
only constant time complexity to calculate the in-
tensity function for every new event in a mutu-
ally exciting Hawkes process. This is achieved
by developing a recursive formulation for mutu-
ally exciting Hawkes process and maintaining an
additional data structure which takes a constant
space. We found considerable improvement in
runtime performance of the Hawkes process ap-
plied to the sequential stance classification task
on synthetic and real world datasets.
Actions (login required)
 |
View Item |
Altmetric
Altmetric