Abstract:Network data in real-world tends to be error-prone due to incomplete sampling, imperfect measurements, etc.; this in turn results in inaccurate results when performing network analysis or modeling, such as node classification and link prediction, on these flawed networks. In this paper, we aim to reconstruct a reliable network from a flawed, undirected, unweighted network, a process referred to network enhancement. More specifically, network enhancement aims to detect the noisy links that are observed in the network but should not exist in the real world, as well as to predict the missing links that do indeed exist in the real world yet remain unobserved. While some attempts have been made to detect either noisy links or missing links, few of these works have considered unifying these two tasks, even though they are inter-dependent and capable of mutually boosting each others’ performance. In this paper, we therefore propose E-Net, an end-to-end graph neural network model, to leverage the mutual influence of these two tasks in order to achieve both goals more effectively. On one hand, detecting noisy links can benefit the performance of missing link prediction, while on the other hand, predicting missing links can provide indirect supervision for detecting noisy link detection when the labels of these noisy links are unavailable. Moreover, by proposing a subgraph extraction mechanism based on random walk with restart, the model can be scaled up to large networks and is able to preserve the local and global structural characteristics. The experimental results on several types of large networks demonstrate that the proposed model obtains an improvement of 10.7 percent on average in terms of F1 for predicting missing links, along with an average of 3.7 percent improvement in terms of precision for detecting noisy links compared with the state-of-the-art baselines.

Robust Estimation from Multiple Graphs under Gross Error Contamination

Robustness of Deep Learning Models on Graphs: A Survey

Unsupervised Adversarially-Robust Representation Learning on Graphs

Robust Network Enhancement from Flawed Networks.

Robust Graph Neural Networks via Unbiased Aggregation

Robust Graph Representation Learning for Local Corruption Recovery

Robust Graph Learning From Noisy Data

Robust estimation with Lasso when outputs are adversarially contaminated

Robust W-GAN-Based Estimation Under Wasserstein Contamination

Robust Causal Graph Representation Learning Against Confounding Effects

A NEW PERSPECTIVE ON ROBUST M-ESTIMATION: FINITE SAMPLE THEORY AND APPLICATIONS TO DEPENDENCE-ADJUSTED MULTIPLE TESTING

Tractable and Near-Optimal Adversarial Algorithms for Robust Estimation in Contaminated Gaussian Models

Optimal Robust Estimation under Local and Global Corruptions: Stronger Adversary and Smaller Error

Transfer Learning for Latent Variable Network Models

Multivariate Regression with Grossly Corrupted Observations: A Robust Approach and its Applications

Statistical inference on errorfully observed graphs

Towards Reliable Link Prediction with Robust Graph Information Bottleneck

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

Robust Sparse Estimation for Gaussians with Optimal Error under Huber Contamination

L1-Penalized Censored Gaussian Graphical Model

Surrogate-Assisted Robust Optimization of Large-Scale Networks Based on Graph Embedding