Abstract:Controllability of complex networks refers to the ability of a networked system to steer from any initial state to any desired state in a finite time under feasible control inputs. Its robustness reflects the ability of a network to maintain controllability under various attacks, which records a series of controllability values under continuous attacks and can be directly determined by attack simulations. However, such simulation experiments are very time-consuming. It is necessary to design a model to replace simulation experiments to learn network controllability robustness with high precision and high speed, which is the objective of Network Controllability Robustness Learning. In this paper, given the superiority of spatial graph neural networks in generating network embedding, a novel controllability robustness learning approach via spatial graph neural networks (CRL-SGNN) is proposed by taking into account both generalization and high precision. In this scheme, controllability robustness related complex network representations that contain network topology information and degree-based node attribute information are first generated through the introduced graph convolutional layer and then sent to the prediction module consisting of multiple branches. A large number of experiments show that: 1) CRL-SGNN can overcome the issue of different distributions between training and test sets with satisfactory performance, implying strong generalization ability of CRL-SGNN; 2) Compared to the cutting-edge methods, CRL-SGNN can obtain superior results in less time; and 3) A highlight which cannot be ignored is that the embedding generation module of CRL-SGNN can serve as a transferable feature learning module to deal with complex networks of any size with low prediction error and high efficiency.

GCRL: A Graph Neural Network Framework for Network Connectivity Robustness Learning

A Graph Transformer-Driven Approach for Network Robustness Learning

Network Controllability Robustness Learning Via Spatial Graph Neural Networks

Robust graph learning with graph convolutional network

Robust Graph Convolutional Networks Against Adversarial Attacks

RGLN: ROBUST RESIDUAL GRAPH LEARNING NETWORKS VIA SIMILARITY-PRESERVING MAPPING ON GRAPHS

Not All Low-Pass Filters Are Robust in Graph Convolutional Networks

Post-Hoc Robustness Enhancement in Graph Neural Networks with Conditional Random Fields

From Decoupling to Reconstruction: A Robust Graph Neural Network Against Topology Attacks.

Machine Learning for Robust Network Design: A New Perspective

Edge Rewiring Goes Neural: Boosting Network Resilience without Rich Features

A Learning Convolutional Neural Network Approach for Network Robustness Prediction

Graph Convolutional Network with elastic topology

SMGRL: Scalable Multi-resolution Graph Representation Learning

Robust Graph Data Learning via Latent Graph Convolutional Representation

Exploring Structure-Adaptive Graph Learning for Robust Semi-Supervised Classification

Certifiably Robust Graph Contrastive Learning

ALG: Fast and Accurate Active Learning Framework for Graph Convolutional Networks.

Turning Strengths into Weaknesses: A Certified Robustness Inspired Attack Framework against Graph Neural Networks

Learning Discrete Adaptive Receptive Fields for Graph Convolutional Networks

Robust graph neural networks with Dirichlet regularization and residual connection