Adrian Arnaiz Rodriguez

7 Followers
19 Following
8 Posts
► PhD Candidate in Algorithmic Fairness in Machine Leaning and GNNs @ ELLIS Alicante (Spain)
► Alumni UBU - UOC
Webpagehttps://adrian-arnaiz.netlify.app/
Scholarhttps://scholar.google.com/citations?hl=es&user=M50V8k4AAAAJ
Twitterhttps://twitter.com/arnaiztech

"DiffWire: Inductive Graph Rewiring via the Lovász bound" has been accepted at the
#LearningOnGraphs conference!

Check this out if you are curious about Graph Rewiring using GNNs. Rewiring is one important technique to overcome current GNN limitations. We propose a theoretical framework for rewiring that is principled, differentiable, inductive, and parameter-free

📄Paper: https://openreview.net/forum?id=IXvfIex0mX6f

🤖Code: https://github.com/AdrianArnaiz/DiffWire

DiffWire: Inductive Graph Rewiring via the Lovász Bound

We propose DiffWire, a unifying GNN rewiring method that is parameter-free and differentiable.

OpenReview

📢 I'm very excited to announce that our tutorial on "Graph Rewiring: from Theory to Applications in Fairness" is accepted in #LearningOnGraphs #LoG conference (Dec 11th) 😄 Francisco Escolano and me will be presenting 2h of theory and hands on content!

But that's not all, we will run a amazing panel discussion. It wil be moderated by Nuria Oliver with the participation of Marinka Zitnik, Petar Velickovic, Francesco Fabbri and Francesco Di Giovanni.

More info (schedule, webpage...) soon 📢