Skip to content

Commit

Permalink
Add PNA to leaderboards (graphdeeplearning#34)
Browse files Browse the repository at this point in the history
  • Loading branch information
@vijaydwivedi75
vijaydwivedi75 authored Jul 20, 2020
1 parent e1cb3a0 commit 03da731
Showing 1 changed file with 34 additions and 25 deletions.
59 changes: 34 additions & 25 deletions docs/07_leaderboards.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -52,15 +52,18 @@ The leaderboard includes the best performing GNN models on each datasets, _in or

|Rank|Model | #Params | Test MAE ± s.d. | Links |
|----| ---------- |------------:| :--------:|:-------:|
|1|GatedGCN-PE|505011 |0.214 ± 0.006|[Paper](https://bit.ly/gatedgcn-pe-paper) |
|2|GatedGCN-E|504309| 0.282 ± 0.015|[Paper](https://bit.ly/gatedgcn-pe-paper) |
|3|MoNet|504013 |0.292 ± 0.006|[Paper](https://bit.ly/monet-paper) |
|4|3WLGNN-E|507603|0.303 ± 0.068|[Paper](https://bit.ly/3wlgnn-paper) |
|5|RingGNN-E|527283| 0.353 ± 0.019|[Paper](https://bit.ly/ring-gnn-paper) |
|6|GCN|505079| 0.367 ± 0.011|[Paper](https://bit.ly/gcn-paper) |
|7|GAT|531345|0.384 ± 0.007|[Paper](https://bit.ly/gat-paper) |
|8|GraphSage|505341 |0.398 ± 0.002|[Paper](https://bit.ly/graphsage-paper) |
|9|GIN|509549| 0.526 ± 0.051|[Paper](https://bit.ly/gin-paper) |
|1|PNA|387155 |0.142 ± 0.010|[Paper](https://bit.ly/pna-paper), [Code](https://bit.ly/pna-code) |
|2|MPNN (sum)|480805 |0.145 ± 0.007|[Paper](https://bit.ly/pna-paper), [Code](https://bit.ly/pna-code) |
|3|GatedGCN-PE|505011 |0.214 ± 0.006|[Paper](https://bit.ly/gatedgcn-pe-paper) |
|4|MPNN (max)|480805 |0.252 ± 0.009|[Paper](https://bit.ly/pna-paper), [Code](https://bit.ly/pna-code) |
|5|GatedGCN-E|504309| 0.282 ± 0.015|[Paper](https://bit.ly/gatedgcn-pe-paper) |
|6|MoNet|504013 |0.292 ± 0.006|[Paper](https://bit.ly/monet-paper) |
|7|3WLGNN-E|507603|0.303 ± 0.068|[Paper](https://bit.ly/3wlgnn-paper) |
|8|RingGNN-E|527283| 0.353 ± 0.019|[Paper](https://bit.ly/ring-gnn-paper) |
|9|GCN|505079| 0.367 ± 0.011|[Paper](https://bit.ly/gcn-paper) |
|10|GAT|531345|0.384 ± 0.007|[Paper](https://bit.ly/gat-paper) |
|11|GraphSage|505341 |0.398 ± 0.002|[Paper](https://bit.ly/graphsage-paper) |
|12|GIN|509549| 0.526 ± 0.051|[Paper](https://bit.ly/gin-paper) |



Expand All @@ -72,14 +75,17 @@ The leaderboard includes the best performing GNN models on each datasets, _in or

|Rank|Model | #Params | Test Acc ± s.d. | Links |
|----| ---------- |------------:| :--------:|:-------:|
|1|GatedGCN|104217 |97.340 ± 0.143|[Paper](https://bit.ly/gatedgcn-paper) |
|2|GraphSage|104337 |97.312 ± 0.097|[Paper](https://bit.ly/graphsage-paper) |
|3|GIN|105434 |96.485 ± 0.252|[Paper](https://bit.ly/gin-paper) |
|4|GAT|110400| 95.535 ± 0.205|[Paper](https://bit.ly/gat-paper) |
|5|3WLGNN|108024 |95.075 ± 0.961|[Paper](https://bit.ly/3wlgnn-paper) |
|6|MoNet|104049 |90.805 ± 0.032|[Paper](https://bit.ly/monet-paper) |
|7|GCN|101365 |90.705 ± 0.218|[Paper](https://bit.ly/gcn-paper) |
|8|RingGNN|105398| 11.350 ± 0.000|[Paper](https://bit.ly/ring-gnn-paper) |
|1|PNA|119812 |97.940 ± 0.120|[Paper](https://bit.ly/pna-paper), [Code](https://bit.ly/pna-code) |
|2|MPNN (max)|109057 |97.690 ± 0.220|[Paper](https://bit.ly/pna-paper), [Code](https://bit.ly/pna-code) |
|3|GatedGCN|104217 |97.340 ± 0.143|[Paper](https://bit.ly/gatedgcn-paper) |
|4|GraphSage|104337 |97.312 ± 0.097|[Paper](https://bit.ly/graphsage-paper) |
|5|MPNN (sum)|109057 |96.900 ± 0.150|[Paper](https://bit.ly/pna-paper), [Code](https://bit.ly/pna-code) |
|6|GIN|105434 |96.485 ± 0.252|[Paper](https://bit.ly/gin-paper) |
|7|GAT|110400| 95.535 ± 0.205|[Paper](https://bit.ly/gat-paper) |
|8|3WLGNN|108024 |95.075 ± 0.961|[Paper](https://bit.ly/3wlgnn-paper) |
|9|MoNet|104049 |90.805 ± 0.032|[Paper](https://bit.ly/monet-paper) |
|10|GCN|101365 |90.705 ± 0.218|[Paper](https://bit.ly/gcn-paper) |
|11|RingGNN|105398| 11.350 ± 0.000|[Paper](https://bit.ly/ring-gnn-paper) |


**Models with configs having 500k trainable parameters for 3WLGNN and RingGNN**
Expand All @@ -100,14 +106,17 @@ The leaderboard includes the best performing GNN models on each datasets, _in or

|Rank|Model | #Params | Test Acc ± s.d. | Links |
|----| ---------- |------------:| :--------:|:-------:|
|1|GatedGCN|104357|67.312 ± 0.311|[Paper](https://bit.ly/gatedgcn-paper) |
|2|GraphSage|104517|65.767 ± 0.308|[Paper](https://bit.ly/graphsage-paper) |
|3|GAT|110704|64.223 ± 0.455|[Paper](https://bit.ly/gat-paper) |
|4|3WLGNN|108516|59.175 ± 1.593|[Paper](https://bit.ly/3wlgnn-paper) |
|5|GCN|101657|55.710 ± 0.381|[Paper](https://bit.ly/gcn-paper) |
|6|GIN|105654|55.255 ± 1.527|[Paper](https://bit.ly/gin-paper) |
|7|MoNet|104229|54.655 ± 0.518|[Paper](https://bit.ly/monet-paper) |
|8|RingGNN|105165|19.300 ± 16.108|[Paper](https://bit.ly/ring-gnn-paper) |
|1|MPNN (max)|109277 |70.860 ± 0.270 |[Paper](https://bit.ly/pna-paper), [Code](https://bit.ly/pna-code) |
|2|PNA|113472 |70.350 ± 0.630 |[Paper](https://bit.ly/pna-paper), [Code](https://bit.ly/pna-code) |
|3|GatedGCN|104357|67.312 ± 0.311|[Paper](https://bit.ly/gatedgcn-paper) |
|4|GraphSage|104517|65.767 ± 0.308|[Paper](https://bit.ly/graphsage-paper) |
|5|MPNN (sum)|109277 |65.610 ± 0.300 |[Paper](https://bit.ly/pna-paper), [Code](https://bit.ly/pna-code) |
|6|GAT|110704|64.223 ± 0.455|[Paper](https://bit.ly/gat-paper) |
|7|3WLGNN|108516|59.175 ± 1.593|[Paper](https://bit.ly/3wlgnn-paper) |
|8|GCN|101657|55.710 ± 0.381|[Paper](https://bit.ly/gcn-paper) |
|9|GIN|105654|55.255 ± 1.527|[Paper](https://bit.ly/gin-paper) |
|10|MoNet|104229|54.655 ± 0.518|[Paper](https://bit.ly/monet-paper) |
|11|RingGNN|105165|19.300 ± 16.108|[Paper](https://bit.ly/ring-gnn-paper) |


**Models with configs having 500k trainable parameters for 3WLGNN and RingGNN**
Expand Down

0 comments on commit 03da731

Please sign in to comment.