This application program estimates both extrinsic environmental changes and intrinsic chain-reaction mechanisms from a single series of event-occurrence times, base on the "Hawkes decoder" we have developed in our paper.

This web-application program was developed by Hidetaka Manabe. An original program was formulated by Shinsuke Koyama. The analysis was directed by Shigeru Shinomoto.

Here, α is the reproduction ratio representing the average number of events induced by a single event, and γ is a hyperparameter that specifies the smoothness or flatness in the change of exogenous factors. These parameters will be automatically optimized by the Empirical Bayes method we developed in our paper. The timescale of self-excitation kernel τ can be changed by filling in the box.

[Reference]
S. Koyama and S. Shinomoto,
The statistical physics of discovering exogenous and endogenous factors in a chain of events.
Physical Review Research (2020) 2:043358 [open access]

Version 2.1: 2021/01/12.