MaxCutPool: differentiable feature-aware Maxcut for pooling in graph neural networks

Carlo Abate, Filippo Maria Bianchi
9/8/2024
cs.LG

No official code found. Best implementation: NGMLGroup/MaxCutPool (8 stars, MIT, last commit 2025-03-08).

Abstract

We propose a novel approach to compute the MAXCUT in attributed graphs, i.e., graphs with features associated with nodes and edges. Our approach works well on any kind of graph topology and can find solutions that jointly optimize the MAXCUT along with other objectives. Based on the obtained MAXCUT partition, we implement a hierarchical graph pooling layer for Graph Neural Networks, which is sparse, trainable end-to-end, and particularly suitable for downstream tasks on heterophilic graphs.

View on arXivView PDF

Code Implementations(1)

NGMLGroup/MaxCutPoolCommunity100%
MIT

Official repository of the paper "MaxCutPool: differentiable feature-aware Maxcut for pooling in graph neural networks" presented at ICLR 2025.

81PythonAug 26, 20241 years agoMIT
graph-classificationgraph-coarseninggraph-neural-networksgraph-poolingmaxcut+2 more

Similar Papers

Maximal Independent Sets for Pooling in Graph Neural Networks

81% match2023

ENADPool: The Edge-Node Attention-based Differentiable Pooling for Graph Neural Networks

81% match2024

Geometry-Aware Edge Pooling for Graph Neural Networks

80% match1 repo72025

Edge-Based Graph Component Pooling

80% match2024

A Benchmark for Maximum Cut: Towards Standardization of the Evaluation of Learned Heuristics for Combinatorial Optimization

80% match1 repo92024

Torch Geometric Pool: the Pytorch library for pooling in Graph Neural Networks

79% match1 repo612025

Cite this paper

@article{abate2024maxcutpool,
  title  = {MaxCutPool: differentiable feature-aware Maxcut for pooling in graph neural networks},
  author = {Carlo Abate and Filippo Maria Bianchi},
  year   = {2024},
  eprint = {2409.05100},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2409.05100v3},
  journal = {ICLR 2024 2024}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2409.05100.svg)](https://codeofpaper.com/paper/2409.05100)

Discussion