Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
lance123123 a4fcd955f1 | 11 months ago | |
---|---|---|
model_files | 11 months ago | |
model_results/DGraphFin | 1 year ago | |
models | 11 months ago | |
output | 1 year ago | |
utils | 11 months ago | |
.gitignore | 1 year ago | |
README.md | 1 year ago | |
basic.yaml | 1 year ago | |
batch.py | 11 months ago | |
gear_gnn.py | 1 year ago | |
gnn.py | 1 year ago | |
gnn_cs.py | 1 year ago | |
gnn_mini_batch.py | 1 year ago | |
gnn_new.py | 1 year ago | |
infer_mini.py | 11 months ago | |
train_mini.py | 11 months ago |
This repo provides a collection of baselines for DGraphFin dataset. Please download the dataset from the DGraph web and place it under the folder './dataset/DGraphFin/raw'.
model_results
: store the .csv file which include the final result for each modelmodels
: model file included by gnn.pyutils
: include the preprocess file / dataset loader / logger / Evaluatoroutput
: the log file to run each modelPython environment:
GPU environment:
python gear_gnn.py --model gear --dataset DGraphFin --epochs 200 --runs 10 --device 0
python gnn.py --model sign --dataset DGraphFin --epochs 200 --runs 10 --device 0
python gnn.py --model mlp --dataset DGraphFin --epochs 200 --runs 10 --device 0
python gnn.py --model gcn --dataset DGraphFin --epochs 200 --runs 10 --device 0
python gnn.py --model sage --dataset DGraphFin --epochs 200 --runs 10 --device 0
python gnn_mini_batch.py --model sage_neighsampler --dataset DGraphFin --epochs 200 --runs 10 --device 0
python gnn_mini_batch.py --model gat_neighsampler --dataset DGraphFin --epochs 200 --runs 10 --device 0
python gnn_mini_batch.py --model gatv2_neighsampler --dataset DGraphFin --epochs 200 --runs 10 --device 0
Performance on DGraphFin(10 runs) (%):
(ranked by test AUC )
rk | Methods | Train AUC | Valid AUC | Test AUC |
---|---|---|---|---|
1 | GEARSAGE | 84.7251 ± 0.0776 | 83.3331 ± 0.0747 | 84.1887 ± 0.0565 |
2 | GraphSAGE (NeighborSampler) | 78.6245 ± 0.1391 | 76.8072 ± 0.08 | 77.6441 ± 0.1343 |
3 | SIGN | 77.2373 ± 0.2803 | 75.5652 ± 0.1840 | 76.9460 ± 0.3002 |
4 | GraphSAGE | 76.7854 ± 0.1881 | 75.4739 ± 0.1894 | 76.2051 ± 0.2010 |
5 | GATv2 (NeighborSampler) | 76.3698 ± 0.7377 | 74.7529 ± 0.788 | 75.7034 ± 0.6571 |
6 | GAT (NeighborSampler) | 74.2509 ± 0.3803 | 72.5287 ± 0.2654 | 73.6141 ± 0.3018 |
7 | MLP | 72.1234 ± 0.0912 | 71.2699 ± 0.0924 | 71.8815 ± 0.0858 |
8 | GCN | 71.0831 ± 0.3224 | 70.7958 ± 0.3028 | 70.7996 ± 0.2721 |
Performance on DGraphFin(10 runs):
Methods | Train AUC | Valid AUC | Test AUC |
---|---|---|---|
MLP | 0.7221 ± 0.0014 | 0.7135 ± 0.0010 | 0.7192 ± 0.0009 |
GCN | 0.7108 ± 0.0027 | 0.7078 ± 0.0027 | 0.7078 ± 0.0023 |
GraphSAGE | 0.7682 ± 0.0014 | 0.7548 ± 0.0013 | 0.7621 ± 0.0017 |
GraphSAGE (NeighborSampler) | 0.7845 ± 0.0013 | 0.7674 ± 0.0005 | 0.7761 ± 0.0018 |
GAT (NeighborSampler) | 0.7396 ± 0.0018 | 0.7233 ± 0.0012 | 0.7333 ± 0.0024 |
GATv2 (NeighborSampler) | 0.7698 ± 0.0083 | 0.7526 ± 0.0089 | 0.7624 ± 0.0081 |
No Description
Python CSV
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》