Deep Graph Clustering via Dual Correlation Reduction

Authors: Yue Liu, Wenxuan Tu, Sihang Zhou, Xinwang Liu, Linxuan Song, Xihong Yang, En Zhu7603-7611

AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experimental results on six benchmark datasets demonstrate the effectiveness of the proposed DCRN against the existing state-of-the-art methods.
Researcher Affiliation Academia 1College of Computer, National University of Defense Technology, Changsha, China 2College of Intelligence Science and Technology, National University of Defense Technology, Changsha, China
Pseudocode Yes Algorithm 1: Dual Correlation Reduction Network
Open Source Code Yes The code of DCRN is available at https://github.com/yueliu1999/DCRN and a collection (papers, codes and, datasets) of deep graph clustering is shared at https://github.com/yueliu1999/Awesome-Deep Graph-Clustering on Github.
Open Datasets Yes To evaluate the effectiveness of the proposed method, we conduct extensive experiments on six widely-used datasets, including DBLP, CITE, ACM(Bo et al. 2020), AMAP, PUBMED, and CORAFULL(Shchur et al. 2018).
Dataset Splits No The paper describes the training procedure (pre-training, integrating subnetworks, training the whole network for certain epochs) but does not provide specific details on train/validation/test splits like percentages, sample counts, or explicit references to predefined splits.
Hardware Specification Yes The proposed DCRN is implemented with a NVIDIA 3090 GPU on Py Torch platform.
Software Dependencies No The paper mentions 'Py Torch platform' but does not provide a specific version number for PyTorch or any other software dependencies.
Experiment Setup Yes The learning rate is set to 1e-3 for AMAP, 1e-4 for DBLP, 5e-5 for ACM, 1e-5 for CITE, PUBMED, and CORAFULL, respectively. The hyper-parameters α is set to 0.1 for PUBMED and 0.2 for other datasets. Moreover, we set λ and γ to 10 and 1e3, respectively. K in Eq. 6 is set to the cluster number C.