xnuohz / arma-dgl

A DGL implementation of "Graph Neural Networks with convolutional ARMA filters". (PAMI 2021)

DGL Implementation of ARMA

This DGL example implements the GNN model proposed in the paper Graph Neural Networks with convolutional ARMA filters. For the original implementation, see here.

Contributor: xnuohz

Requirements

The codebase is implemented in Python 3.6. For version requirement of packages, see below.

dgl 0.6a210202
numpy 1.19.5
networkx 2.5
scikit-learn 0.24.1
tqdm 4.56.0
torch 1.7.0s

The graph datasets used in this example

Node Classification

The DGL's built-in Cora, Pubmed, Citeseer and PPI datasets. Dataset summary:

Dataset	#Nodes	#Edges	#Feats	#Classes	#Train Nodes	#Val Nodes	#Test Nodes
Cora	2,708	10,556	1,433	7(single label)	140	500	1000
Citeseer	3,327	9,228	3,703	6(single label)	120	500	1000
Pubmed	19,717	88,651	500	3(single label)	60	500	1000
PPI	56,944	818,716	50	121(multi label)	44906(20 graphs)	6514(2 graphs)	5524(2 graphs)

Graph Classification

Dataset	#Samples	#Classes	#Avg. nodes	#Avg. edges	#Node attr.	Node labels
Enzymes	600	6	32.63	62.14	18	no
Proteins	1,113	2	39.06	72.82	1	no
D&D	1,178	2	284.32	715.66	-	yes
MUTAG	188	2	17.93	19.79	-	yes

Usage

Dataset options

--dataset          str     The graph dataset name.             Default is 'Cora'.

GPU options

--gpu              int     GPU index.                          Default is -1, using CPU.

Model options

--epochs           int     Number of training epochs.          Default is 2000.
--early-stopping   int     Early stopping rounds.              Default is 100.
--lr               float   Adam optimizer learning rate.       Default is 0.01.
--lamb             float   L2 regularization coefficient.      Default is 0.0005.
--hid-dim          int     Hidden layer dimensionalities.      Default is 16.
--num-stacks       int     Number of K.                        Default is 2.
--num-layers       int     Number of T.                        Default is 1.
--dropout          float   Dropout applied at all layers.      Default is 0.75.

Examples

The following commands learn a neural network and predict on the test set. Train an ARMA model which follows the original hyperparameters on different datasets.

# Cora:
python citation.py --gpu 0

# Citeseer:
python citation.py --gpu 0 --dataset Citeseer --num-stacks 3

# Pubmed:
python citation.py --gpu 0 --dataset Pubmed --dropout 0.25 --num-stacks 1

# PPI:
python ppi.py --gpu 0

# Enzymes
python tu.py --gpu 0

# Proteins
python tu.py --gpu 0 --dataset PROTEINS --num-stacks 4 --num-layers 4

# D&D
python tu.py --gpu 0 --dataset DD --dropout 0 --num-stacks 4 --num-layers 4

# MUTAG
python tu.py --gpu 0 --dataset MUTAG --dropout 0 --num-stacks 4 --num-layers 4

Performance

Node Classification

Dataset	Cora	Citeseer	Pubmed	PPI
Metrics(Table 1.Node classification accuracy/f1)	83.4±0.6	72.5±0.4	78.9±0.3	90.5±0.3
Metrics(PyG)	82.3±0.5	70.9±1.1	78.3±0.8	-
Metrics(DGL)	80.9±0.6	71.6±0.8	75.0±4.2	73.2±0.1

Graph Classification

Dataset	Enzymes	Proteins	D&D	MUTAG
Metrics(Table 3.Graph classification accuracy.)	60.6±7.2	73.7±3.4	77.6±2.7	91.5±4.2
Metrics(DGL)	98.2±2.2	100.0±0.0	100.0±0.0	13.0±26