Learning Hawkes Processes from a handful of events

Authors: Farnood Salehi, William Trouleau, Matthias Grossglauser, Patrick Thiran

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on both synthetic and real datasets show that our approach significantly outperforms state-of-the-art methods under short observation sequences.
Researcher Affiliation Academia Farnood Salehi EPFL farnood.salehi@epfl.ch William Trouleau EPFL william.trouleau@epfl.ch Matthias Grossglauser EPFL matthias.grossglauser@epfl.ch Patrick Thiran EPFL patrick.thiran@epfl.ch
Pseudocode Yes Algorithm 1 summarizes the proposed variational EM approach.
Open Source Code Yes Source code is available publicly.
Open Datasets Yes 1. Epidemics. This dataset contains records of infection of individuals, along with their correspond-ing district of residence, during the last Ebola epidemic in 2014-2015 [15]. 2. Stock market. This dataset contains the stock prices of 12 high-tech companies sampled every 2 minutes on the New York Stock Exchange for 20 days in April 2008 [13]. 3. Enron email. This dataset contains emails between employees of Enron from the Enron corpus.
Dataset Splits No To do so, we use the first 70% events as training set, and we compute the held-out averaged log-likelihood on the remaining 30%. This indicates a train/test split, but no explicit validation split or cross-validation is mentioned.
Hardware Specification No The paper does not explicitly specify any hardware details such as GPU/CPU models, processors, or memory used for experiments.
Software Dependencies No The paper does not provide specific software names with version numbers for its dependencies.
Experiment Setup Yes For reproducibility, a detailed description of the experimental setup is provided in Appendix E.