Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
Theheavens 299b0523f9 | 2 years ago | |
---|---|---|
.. | ||
README.md | 2 years ago |
Paper: Heterogeneous Graph Neural Network
Code from author: https://github.com/chuxuzhang/KDD2019_HetGNN
Clone the Openhgnn-DGL
python main.py -m HetGNN -t node_classification -d academic4HetGNN -g 0
If you do not have gpu, set -g -1.
Node classification | Macro-F1 | Micro-F1 |
---|---|---|
paper | 0.978 | 0.979 |
OpenHGNN | 0.9701 | 0.9705 |
A-II(type-1) authors link prediction | AUC | F1 |
---|---|---|
paper | 0.717 | 0.669 |
OpenHGNN | 0.7235 | 0.7205 |
We process the Academic II dataset given by HetGNN. It saved as dgl.heterograph and can be loaded by dgl.load_graphs
You can download the dataset by
wget https://s3.cn-north-1.amazonaws.com.cn/dgl-data/dataset/academic4HetGNN.zip
Or run the code mentioned above and it will download automaticlly.
You can modify the parameters in openhgnn/config.ini
batch_size # the batch_size of nodes sampled in SkipGramBatchSampler
window_size # window size for relation extration
num_workers # how many subprocesses to use for data loading. ''0'' means that the data will be loaded in the main process.
batches_per_epoch # the batches trained in a epoch
# the three parameters decided in what nodes aggregated.[used in building hetgnn_graph]
rw_length # length of each walk
rw_walks # window size for relation extration
rwr_prob # the probability of restart
#train epoch and early stop patience
max_epoch
patience
mini_batch_flag = True # we only support the mini-batch training.
Best config can be found in best_config
Tianyu Zhao[GAMMA LAB]
Submit an issue or email to tyzhao@bupt.edu.cn.
OpenHGNN是由北邮GAMMA Lab开发的基于PyTorch和DGL的开源异质图神经网络工具包。
Python Markdown Shell
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》