Effective Decoding in Graph Auto-Encoder Using Triadic Closure

Authors: Han Shi, Haozheng Fan, James T. Kwok906-913

AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on link prediction, node clustering and graph generation show that the use of triads leads to more accurate prediction, clustering and better preservation of the graph characteristics.
Researcher Affiliation Collaboration 1Department of Computer Science and Engineering Hong Kong University of Science and Technology, Hong Kong 2Amazon
Pseudocode Yes Algorithm 1 Training the triad variational graph autoencoder (TVGA) and triad graph auto-encoder (TGA) using SGD.
Open Source Code No The paper does not provide any explicit statement or link regarding the availability of its source code.
Open Datasets Yes Experiments are performed on three standard benchmark citation graph data sets1 (Sen et al. 2008): Cora, Citeseer, and Pubmed (Table 1). 1http://www.cs.umd.edu/ sen/lbc-proj/LBC.html
Dataset Splits Yes In this experiment, 85% of the edges and non-edges (unconnected nodes) from each graph are randomly selected to form the training set, another 10% is used as the validation set, and the remaining 5% as testing set.
Hardware Specification No The paper does not provide specific details about the hardware (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper mentions 'Adam (Kingma and Ba 2014) is the optimizer' but does not specify version numbers for any software dependencies or libraries.
Experiment Setup Yes The proposed algorithm uses a mini-batch size of 5,000. Adam (Kingma and Ba 2014) is the optimizer, with a learning rate of 0.0005. Both the hidden layer and embedding layer of the encoder have 32 hidden units. The convolution layer in the triad decoder has 4 filters (i.e., the dimension of ztriplet is 1 32 4).