Adversarial Training Methods for Network Embedding

Abstract

Network Embedding is the task of learning continuous node representations for networks, which has been shown effective in a variety of tasks such as link prediction and node classification. Most of existing works aim to preserve different network structures and properties in low-dimensional embedding vectors, while neglecting the existence of noisy information in many real-world networks and the overfitting issue in the embedding learning process. Most recently, generative adversarial networks (GANs) based regularization methods are exploited to regularize embedding learning process, which can encourage a global smoothness of embedding vectors. These methods have very complicated architecture and suffer from the well-recognized non-convergence problem of GANs. In this paper, we aim to introduce a more succinct and effective local regularization method, namely adversarial training, to network embedding so as to achieve model robustness and better generalization performance. Firstly, the adversarial training method is applied by defining adversarial perturbations in the embedding space with an adaptive L 2 norm constraint that depends on the connectivity pattern of node pairs. Though effective as a regularizer, it suffers from the interpretability issue which may hinder its application in certain real-world scenarios. To improve this strategy, we further propose an interpretable adversarial training method by enforcing the reconstruction of the adversarial examples in the discrete graph domain. These two regularization methods can be applied to many existing embedding models, and we take DeepWalk as the base model for illustration in the paper. Empirical evaluations in both link prediction and node classification demonstrate the effectiveness of the proposed methods. The source code is available online 1 .

CommunityGAN: Community Detection with Generative Adversarial Nets

“…GraphGAN [25] proposes a unified adversarial learning framework, which naturally captures structural information from graphs to learn the graph representation. ANE [5] utilizes GAN as a regularizer for learning stable and robust feature extractor. However, all of the above algorithms focus on general graph representation learning.…”

Graph-Based Method for Anomaly Detection in Functional Brain Network using Variational Autoencoder

“…Pan and et al in [13] proposed an adversarial training scheme to regularize and enforce the latent code to match a prior distribution with a graph convolutional Autoencoder. Makhzani in [14] showed an adversarial Autoencoder to learn the latent embedding by merging the adversarial mechanism into Autoencoder for general data but Dai and et al [15] applied the adversarial procedure for the graph embedding. Also in [12] used an encoder with edge condition convolution (ECC) [16] and condition both encoder and decoder which associated with each of the input graph, this method is useful only for generation small graphs.…”

An Optimized Network Representation Learning Algorithm Using Multi-Relational Data

“…There exist some network representation approaches for special network structures, such as DynamicTriad [25], DepthLGP [26], and DHNE [27], etc. In addition, motivated by generative adversarial networks (GANs) [28], some network representation learning algorithms introduce GANs to optimize the learning procedure, such as ANE [29], GraphGAN [30], and NetGan [31].…”