List your company
ResearchFormerly WizWand

Graph Classification on MUV MoleculeNet random (train val test)

0.798ROC-AUCDuvenaud et al.
0.593120.646310.69950.75269Mar 6, 2019
Updated 8mo ago

Evaluation results

MethodLinks
Duvenaud et al.Relational Pooling=false, Base Model (f)=Duvenaud et al. GNN2019.030.798
RP-DuvenaudRelational Pooling=true, Base Model (f)=Duvenaud et al. GNN2019.030.794
k-ary RP (k=20)Relational Pooling=true, k (subgraph size)=20, Base Model (f)=Duvenaud et al. GNN2019.030.777
k-ary RP (k=40)Relational Pooling=true, k (subgraph size)=40, Base Model (f)=Duvenaud et al. GNN2019.030.776
k-ary RP (k=30)Relational Pooling=true, k (subgraph size)=30, Base Model (f)=Duvenaud et al. GNN2019.030.776
k-ary RP (k=10)Relational Pooling=true, k (subgraph size)=10, Base Model (f)=Duvenaud et al. GNN2019.030.773
k-ary RP (k=50)Relational Pooling=true, k (subgraph size)=50, Base Model (f)=Duvenaud et al. GNN2019.030.768
RNN-DFSRelational Pooling=true, Base Model (f)=RNN, Orientation (DFS)=true2019.030.648
CNN-DFSRelational Pooling=true, Base Model (f)=CNN, Orientation (DFS)=true2019.030.601