The proposed cross-domain task coordinator (CDTC) for cross-domain few-shot graph learning.

Environment

We used the following Python packages for core development. We tested on Python 3.7.

- pytorch 1.7.0
- torch-geometric  2.0.3
- torch-cluster             1.5.8                   
- torch-scatter             2.0.5                  
- torch-sparse              0.6.8                  
- torch-spline-conv         1.2.0

Datasets

Tox21, SIDER, MUV are public datasets from molecular property prediction benchmarks from SNAP.

Experiments

Example of executing the cross-domain experiments:

python main_ms2t_rl.py --test_dataset $TESTDATA$ --d_names $TRAINDATA-TESTDATA$ --eval_steps 10 --inner_lr $INNER_LR$ --update_step_test  4  --batch_task 4 --seed 5  --naivedim $SOURCE_TASK_NUMBER$  --meta_lr 0.0005  --n_shot_train 10 --n_shot_test 10 --pretrained $PRETRAINED_OR_NOT$ --epochs 2000 --enc_gnn $GNN_TYPE$ --maml --applyrl  --rl_adv $AGENT_LR2$  --step_gin1 10 --step_agent 500