Research map / Graph learning / Graph algorithms and clustering

MMD Graph Kernel: Effective Metric Learning for Graphs via Maximum Mean Discrepancy

ICLR 2024 · Graph learning › Graph algorithms and clustering · approach: Graph embedding and spectral sparsification

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes MMD-based graph kernels (MMD-GK) that apply maximum mean discrepancy to message-passing node representations, plus deep, unsupervised-learnable variants for graph metric learning.

Problem

Learning effective metrics/similarity measures between graphs for graph-level tasks, where existing graph kernels and metrics may be inadequate.

Approach

Computes graph kernels by applying maximum mean discrepancy to node representations obtained via message-passing propagation, and extends this to deep MMD graph kernels that learn kernel and implicit graph features adaptively in an unsupervised manner.

Closest papers by meaning

• Rethinking Kernel Methods for Node Representation Learning on Graphs (NeurIPS 2019 · similarity 0.89)
• Graph Minimum Factor Distance and Its Application to Large-Scale Graph Data Clustering (ICML 2025 · similarity 0.88)
• Hunt For The Unique, Stable, Sparse And Fast Feature Learning On Graphs (NeurIPS 2017 · similarity 0.87)
• Convolutional Kernel Networks for Graph-Structured Data (ICML 2020 · similarity 0.87)
• Graph Classification via Reference Distribution Learning: Theory and Practice (NeurIPS 2024 · similarity 0.86)
• DHAKR: Learning Deep Hierarchical Attention-Based Kernelized Representations for Graph Classification (AAAI 2025 · similarity 0.85)