Explainable AI for analyzing the decision of GNNs at predicting dynamic stability of complex oscillator networks.

Understanding the synchronization of complex oscillator networks is a central question in complex systems research. Recent studies have shown that graph neural networks (GNNs) outperform a wide range of traditional network measures in predicting probabilistic stability in synthetic power grids based on the Kuramoto model. This suggests that analyzing GNN decisions could enhance our understanding of synchronization patterns. We use explainable artificial intelligence (XAI), specifically Layer-wise Relevance Propagation (LRP) and its adaptation for GNNs (GNN-LRP) to analyze these decision processes. Our results indicate that larger neighborhoods beyond direct nodes strongly influence the dynamic behavior, with slightly different patterns for nodes with low stability compared to stable nodes. Aggregating LRP scores provides a nodal measure of stability contribution, correlating with some network measures and suggesting pathways to more stable power grids. However, GNN decision processes appear to be more complex and not only influenced by established node metrics. Our study highlights the potential of GNNs and XAI in understanding synchronization patterns of oscillator networks and emphasizes the need for new XAI methods tailored to this domain.

• Publication year

  2025

• Venue

  Chaos

• Publication date

  2025-11-01

• Fields of study

  Medicine, Computer Science, Engineering

• Identifiers
  DOI 10.1063/5.0278469PMID 41217265
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar, PubMed

• No claims are published for this paper.

• No concepts are published for this paper.

• Predicting Instability in Complex Oscillator Networks: Limitations and Potentials of Network Measures and Machine Learning

  2024cited by this paper
• Explainable Artificial Intelligence (XAI) 2.0: A Manifesto of Open Challenges and Interdisciplinary Research Directions

  2023cited by this paper
• Toward dynamic stability assessment of power grid topologies using graph neural networks.

  2022cited by this paper
• Collective nonlinear dynamics and self-organization in decentralized power grids

  2022cited by this paper
• Towards dynamic stability analysis of sustainable power grids using graph neural networks

  2022influential reference
• Explainable Artificial Intelligence (XAI) techniques for energy and power systems: Review, challenges and opportunities

  2022cited by this paper
• Predicting basin stability of power grids using graph neural networks

  2021cited by this paper
• Higher-Order Explanations of Graph Neural Networks via Relevant Walks

  2020influential reference
• Explainability in Graph Neural Networks: A Taxonomic Survey

  2020cited by this paper
• Graph Representation Learning

  2020cited by this paper
• A Comprehensive Survey on Graph Neural Networks

  2019cited by this paper
• A Survey of Methods for Explaining Black Box Models

  2018cited by this paper
• Topology adaptive graph convolutional networks

  2017cited by this paper
• Neural Message Passing for Quantum Chemistry

  2017cited by this paper
• Deciphering the imprint of topology on nonlinear dynamical network stability

  2016cited by this paper
• Explaining nonlinear classification decisions with deep Taylor decomposition

  2015cited by this paper
• The Kuramoto model in complex networks

  2015cited by this paper
• On Pixel-Wise Explanations for Non-Linear Classifier Decisions by Layer-Wise Relevance Propagation

  2015cited by this paper
• A random growth model for power grids and other spatially embedded infrastructure networks

  2014cited by this paper
• How dead ends undermine power grid stability

  2014cited by this paper
• Cluster synchronization and isolated desynchronization in complex networks with symmetries

  2013cited by this paper
• Second order centrality: Distributed assessment of nodes criticity in complex networks

  2011influential reference
• Communicability betweenness in complex networks

  2009cited by this paper
• Neuronal Oscillations in Cortical Networks

  2004cited by this paper
• Universal behavior of load distribution in scale-free networks.

  2001cited by this paper
• An adaptive model for synchrony in the firefly Pteroptyx malaccae

  1991cited by this paper
• Mechanisms of Sinoatrial Pacemaker Synchronization: A New Hypothesis

  1987cited by this paper
• A Structure Preserving Model for Power System Stability Analysis

  1981cited by this paper
• A set of measures of centrality based upon betweenness

  1977cited by this paper
• Self-entrainment of a population of coupled non-linear oscillators

  1975cited by this paper
• A new status index derived from sociometric analysis

  1953cited by this paper

• No citing papers are available for this paper.