How to get sparse feature importance from GNNExplainer

[BrownEyes1225]

❓ Questions and Help

Hi to everyone!

I am using GNNExplainer in DGL.

I have tried to increase the beta1 and beta 2 to larger values in order to get sparse feature importance vector, but no matter how large the two parameters are, the feature importance are not sparse.

Does ‘sparse’ mean the feature importance should have zeros? If so, why GNNExplainer fail to output sparse results?

Thanks a lot!

[BarclayII]

Both options only make most of the entries go close to zero (like 0.001) by entropy and mean regularization, but not exactly zero. If you want a "truly" sparse explanation you could set a threshold and cut all the masks below it to zero.

[github-actions[bot]]

This issue has been automatically marked as stale due to lack of activity. It will be closed if no further activity occurs. Thank you

[peizhou001]

Hi @BrownEyes1225, I am closing this issue assuming you are happy about our response. Feel free to follow up and reopen the issue if you have more questions with regard to our response.