NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:2805
Title:Mutually Regressive Point Processes

This article proposes a novel continuous-time model for interacting time-event data. The model is able to capture both excitatory and inhibitory interactions. A complete Gibbs sampler is described for posterior inference. The paper is well written. The experimental results are interesting, although the predictive performances do not demonstrate a huge gain in using the proposed model compared to simpler alternatives (even considering the additional experiments provided by the authors in the response).