-
Notifications
You must be signed in to change notification settings - Fork 17
Expander Hypergraph Lifting (Graph to Hypergraph)
This lifting generates an expander (hyper-)graph, more precisely a random Ramanujan graph. It is inspired by recent works on expander graph propagation and expander graph transformers. Expander graphs have favourable, mathematical guarantees w.r.t. node connectivity. E.g., they enable message propagation from any node to any other node in few iterations.
existing lifting from the literature (however GNN literature) | connectivity-based lifting | non-deterministic lifting
Submission by team MIA-UT: Patryk Rygiel (@PatRyg99), Julian Suk (@sukjulian)
From https://github.com/pyt-team/challenge-icml-2024/pull/23
- Defining GCCNs
- Defining backbone models
- Reproducing experiments
-
Graph to Simplicial Complex
-
Graph to Cell Complex
-
Graph to Hypergraph
-
Graph to Combinatorial
-
Pointcloud to Graph
-
Pointcloud to Simplicial
-
Pointcloud to Hypergraph
-
Hypergraph to Simplicial
-
Hypergraph to Combinatorial