1
$\begingroup$

Difference of the model design.

It seems the difference is that GraphSAGE sample the data.

But what is the difference in model architecture.

$\endgroup$

2 Answers 2

1
$\begingroup$

The main novelty of GraphSAGE is a neighborhood sampling step (but this is independent of whether these models are used inductively or transductively). You can think of GraphSAGE as GCN with subsampled neighbors.

In practice, both can be used inductively and transductively.

The title of the GraphSAGE paper ("Inductive representation learning") is unfortunately a bit misleading in that regard. The main benefit of the sampling step of GraphSAGE is scalability (but at the cost of higher variance gradients).

$\endgroup$
0
$\begingroup$

GraphSage provides a solution to address the problem DeepWalk embedding technique. As we know that DeepWalk embedding technique use transudative learning to extract features from a graph. If a node is added in the graph then we gain re-run the algorithm to get embedding of all node. So, DeepWalk besed GNN is not suitable for dynamic graphs where the nodes in the graphs are ever-changing. To address the above-mentioned issue, GraphSage is introduced to learn the node representation in inductive way. Specifically, each node is represented by the aggregation of its neighborhood. Thus, even if a new node unseen during training time appears in the graph, it can still be properly represented by its neighboring nodes.

You can learn more following blog: https://towardsdatascience.com/a-gentle-introduction-to-graph-neural-network-basics-deepwalk-and-graphsage-db5d540d50b3

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.