Dynamic Spiking Graph Neural Networks

Authors: Nan Yin, Mengzhu Wang, Zhenghan Chen, Giulia De Masi, Huan Xiong, Bin Gu

AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on three large-scale real-world dynamic graph datasets validate the effectiveness of Dy-SIGN on dynamic node classification tasks with lower computational costs.
Researcher Affiliation Collaboration Nan Yin1, Mengzhu Wang2, Zhenghan Chen3, Giulia De Masi4, Huan Xiong5,1*, Bin Gu6,1* 1Mohamed bin Zayed University of Artificial Intelligence 2School of Artificial Intelligence, Hebei University of Technology 3Microsoft Corporation 4Technology Innovation Institute 5Harbin Institute of Technology 6Jilin University
Pseudocode Yes Algorithm 1: Learning Algorithm of Dy-SIGN
Open Source Code No The paper does not provide any concrete statement or link regarding the availability of its source code.
Open Datasets Yes We conduct experiments on three large real-world graph datasets, i.e., DBLP (Lu et al. 2019), Tmall (Lu et al. 2019) and Patent (Hall, Jaffe, and Trajtenberg 2001).
Dataset Splits Yes As for the implementation, we follow the same settings with (Li et al. 2023) and report the Macro-F1 and Micro-F1 results under different training ratios (i.e., 40%, 60%, and 80%). Besides, we use 5% for validation.
Hardware Specification No The paper mentions 'under the same experiment environment' regarding memory consumption but does not specify any particular hardware components like GPU or CPU models.
Software Dependencies No The paper does not provide specific version numbers for software dependencies or libraries used in the experiments.
Experiment Setup Yes The hidden dimension of all the methods is set to 128, and the batch size to 1024. The total training epochs are 100 and the learning rate is 0.001.