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.

Using dependence tree model and upward-downward algorithm for network loss tomography

Inference of Link Loss Rates by Explicit Estimation

Fast Parameter Estimation in Loss Tomography for Networks of General Topology

Robust Network Enhancement from Flawed Networks.

A bottom-up inference of loss rate

Simultaneous Topology and Loss Tomography Via a Modified Theme Dictionary Model

A Fast Method To Estimate Loss Rate

Robust Network Tomography: K-Identifiability and Monitor Assignment

Learning-Based Network Boolean Tomography for Identifying Congested Links with Correlations.

The Inference of Link Loss Rates with Internal Monitors.

Loss Tomography from Tree Topologies to General Topologies

Identify Congested Links with Network Tomography under Multipath Routing

Transfer Learning Based High-Speed Link Transient Modeling Method

Lossy link identification based on maximum pseudo likelihood estimation

Taming Both Predictable and Unpredictable Link Failures for Network Tomography.

Network Topology Tomography under Multipath Routing

Fine-Grained Loss Tomography in Dynamic Sensor Networks

DRL-Tomo: a deep reinforcement learning-based approach to augmented data generation for network tomography

General Identifiability Condition for Network Topology Monitoring with Network Tomography

Accurate Per-Link Loss Tomography in Dynamic Sensor Networks

Transient Modeling of High-Speed Links Using Transfer Learning-Based Neural Network Initialization