Learning Fair Graph Representations via Automated Data Augmentations

Authors: Hongyi Ling, Zhimeng Jiang, Youzhi Luo, Shuiwang Ji, Na Zou

ICLR 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results demonstrate that our Graphair consistently outperforms many baselines on multiple node classification datasets in terms of fairness-accuracy trade-off performance.
Researcher Affiliation Academia Hongyi Ling, Zhimeng Jiang, Youzhi Luo, Shuiwang Ji , Na Zou Texas A&M University College Station, TX 77843, USA {hongyiling,zhimengj,yzluo,sji,nzou1}@tamu.edu
Pseudocode Yes We summarize the training algorithm for Graphair and provide the pseudo codes in Algorithm 1.
Open Source Code Yes Our code is publicly available as part of the DIG package (https://github.com/divelab/DIG).
Open Datasets No The paper mentions 'NBA is extended from a Kaggle dataset' and 'Pokec-z and Pokec-n are sampled from a larger social network Pokec' but does not provide specific links, DOIs, repository names, or formal citations with authors and year for direct public access to these datasets.
Dataset Splits Yes we randomly split 10%/10%/80% for training, validating and testing the classifier. ... We randomly split 20%/35%/45% for training, validating and testing the classifier.
Hardware Specification No The paper does not provide specific hardware details such as exact GPU/CPU models, processor types, or memory amounts used for running experiments.
Software Dependencies No The paper mentions using GCN models and Adam optimizer but does not provide specific version numbers for any software dependencies (e.g., Python, PyTorch, TensorFlow, CUDA versions).
Experiment Setup Yes For Graphair, we adopt two-layer GCN models as the adversary model k and augmentation encoder genc, and a three-layer GCN model as the representation encoder f. We use 64 as the hidden dimension in all three models. For the augmentation model, we use an MLP model with 2 layers, the hidden size of 64, and Re LU as the non-linear activation function for MLPA and MLPX. The hyperparameter β is set to 1, and the hyperparameters α, γ and λ are determined with a grid search among {0.1, 1, 10}. ... We train the models for 500 epochs using Adam optimizer with 1 × 10−4 learning rate and 1 × 10−5 weight decay.