Skip to content

Attention-based GNN (AGNN) on Cora

Author: K3-Node Team
Backend: Multi-Backend
Dataset: Cora (Planetoid)
Description: Classify papers in the Cora citation network by topic.

View in Colab   GitHub source

Attention-based GNN (AGNN) on Cora

Classify papers in the Cora citation network by topic. AGNN (Thekumparampil et al., 2018) replaces fully-connected layers by attention-based propagation: each paper averages its neighbors, weighted by how similar they are.

Same model as PyG's examples/agnn.py.

Install K3-Node, then choose a backend: "tensorflow", "torch" or "jax"

!pip install k3-node[examples]
import os
os.environ["KERAS_BACKEND"] = "tensorflow"

Load the data

data holds one graph: node features x, edges edge_index, labels y, and masks marking the training, validation and test nodes.

import keras
from keras import ops
from k3_node.datasets import Planetoid
from k3_node.loader import FullGraphDataset
from k3_node.transforms import NormalizeFeatures

dataset = Planetoid("data/Planetoid", name="Cora", transform=NormalizeFeatures())
data = dataset[0]
print(data)

Define the model

from k3_node.layers import AGNNConv


class AGNN(keras.Model):
    def __init__(self, in_channels, out_channels):
        super().__init__()
        self.dropout = keras.layers.Dropout(0.5)
        self.lin1 = keras.layers.Dense(16, activation="relu")
        self.prop1 = AGNNConv(requires_grad=False)
        self.prop2 = AGNNConv(requires_grad=True)
        self.lin2 = keras.layers.Dense(out_channels)

    def call(self, data, training=False):
        x = self.dropout(data.x, training=training)
        x = self.lin1(x)
        x = self.prop1(x, data.edge_index)
        x = self.prop2(x, data.edge_index)
        x = self.dropout(x, training=training)
        return self.lin2(x)


model = AGNN(dataset.num_features, dataset.num_classes)

Train

FullGraphDataset feeds the whole graph to Keras; mask selects which nodes count in the loss and the accuracy.

model.compile(
    optimizer=keras.optimizers.Adam(learning_rate=0.01, weight_decay=0.0005),
    loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True),
    weighted_metrics=["accuracy"],
)
model.fit(
    FullGraphDataset(data, mask="train_mask"),
    validation_data=FullGraphDataset(data, mask="val_mask"),
    epochs=200,
    verbose=2,
)

Evaluate

loss, accuracy = model.evaluate(FullGraphDataset(data, mask="test_mask"), verbose=0)
print(f"Test accuracy: {accuracy:.4f}")