Initializing Then Refining: A Simple Graph Attribute Imputation Network

Authors: Wenxuan Tu, Sihang Zhou, Xinwang Liu, Yue Liu, Zhiping Cai, En Zhu, Changwang Zhang, Jieren Cheng

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

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on four benchmark datasets verify the superiority of ITR against state-of-the-art methods.
Researcher Affiliation Collaboration Wenxuan Tu1 , Sihang Zhou1 , Xinwang Liu1 , Yue Liu1 , Zhiping Cai1 , En Zhu1 , Changwang Zhang2 and Jieren Cheng3 1National University of Defense Technology, Changsha, China 2Tencent Technology, Shenzhen, China 3Hainan University, Haikou, China
Pseudocode No The paper includes mathematical equations and a framework diagram, but it does not contain a clearly labeled 'Pseudocode' or 'Algorithm' block.
Open Source Code No The paper does not contain any statement or link indicating the availability of open-source code for the described methodology.
Open Datasets Yes We conduct experiments to evaluate the proposed ITR on four benchmark datasets, including Cora [Mc Callum et al., 2000], Citeseer [Sen et al., 2008], Amazon Computer (Amac), and Amazon Photo (Amap) [Shchur et al., 2018].
Dataset Splits Yes Specifically, 1) in the profiling task, we randomly sample 40% nodes with attributes as the training set, and manually mask all attributes of the rest of 10% and 50% nodes (i.e., attribute-missing samples) as the validation set and the test set, respectively. ... 2) in the node classification task, the restored attributes are randomly split into 80% and 20% for training and testing.
Hardware Specification No The paper does not provide specific details about the hardware used, such as GPU or CPU models.
Software Dependencies No The paper mentions using the Adam optimization algorithm and a GCN-based framework but does not specify the versions of any software libraries or dependencies used (e.g., Python, PyTorch, TensorFlow versions).
Experiment Setup Yes We employ a symmetric backbone framework consisting of 4-layer GCNs and optimize it with the Adam optimization algorithm, where we set the learning rate to 1e-3. Moreover, the learning rate, the latent dimension, the dropout rate, and the weight decay are set to 1e-3, 64, 0.5, and 5e-4, respectively.