2月2日論文推薦（附下載地址）
論文題目
Semi-supervised Learning on Graphs with Generative Adversarial Nets
作者
-
Ming Ding
-
Jie Tang
-
Jie Zhang
推薦理由
這是第一篇將生成對抗網路的思想用於圖上的半監督學習任務的工作，達到了state-of-art的效果；除此之外，文章也形象地闡述了生成的樣本如何提升半監督學習的效果並給出博弈論的表示和理論分析。
工作的動機在不同聚類簇之間的density gap裡生成假樣本，讓分類學到的分類函式在分辨真假的同時，阻礙了density gap中的天然連續性。
為了生成density gap中的樣本，文章構造了一種特殊的生成器-判別器的博弈均衡狀態。使得表示層的中心區域成為density gap。
文章在取得state-of-art效果的同時，也分析了一些現象，如被節點判別為假的概率與分類函式的光滑程度之間的關係，很有啟發意義。
摘要
We investigate how generative adversarial nets (GANs) can help semi-supervised learning on graphs. We first provide insights on working principles of adversarial learning over graphs and then present GraphSGAN, a novel approach to semi-supervised learning on graphs. In GraphSGAN, generator and classifier networks play a novel competitive game. At equilibrium, generator generates fake samples in low-density areas between subgraphs. In order to discriminate fake samples from the real, classifier implicitly takes the density property of subgraph into consideration. An efficient adversarial learning algorithm has been developed to improve traditional normalized graph Laplacian regularization with a theoretical guarantee.
Experimental results on several different genres of datasets show that the proposed GraphSGAN significantly outperforms several state-of-the-art methods. GraphSGAN can be also trained using mini-batch, thus enjoys the scalability advantage.
下載地址
https://www.aminer.cn/archive/semi-supervised-learning-on-graphs-with-generative-adversarial-nets/5bdc316717c44a1f58a06fb6