Source code for AAAI2023 "T2-GNN: Graph Neural Networks for Graphs with Incomplete Features and Structure via Teacher-Student Distillation"
- Python3
- NumPy
- SciPy
- PyTorch
- TensorFlow.keras
Before running the code, please unzip the data.zip.
python run.py --dataset texas --Ts 4.0 --topk 10 --lambd 0.8
Please refer to the args.py for more parameters.
The demo code is implemented based on GCN-with-Hinton-Knowledge-Distillation
If you make advantage of T2-GNN in your research, please cite the following in your manuscript:
Cuiying Huo, et al. "T2-GNN: Graph Neural Networks for Graphs with Incomplete Features and Structure via Teacher-Student Distillation." In AAAI. 2023.