arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2112.09752cs.LG
收藏

单跳节点分类的集合Twister

Set Twister for Single-hop Node Classification

Purdue University(普渡大学)

Yangze Zhou, Vinayak Rao, Bruno Ribeiro

摘要:

节点分类是关系学习中的中心任务,当前最先进的技术取决于两个关键原则:(i)预测对于节点的邻居的排序是置换不变的,以及(ii)预测是节点的$r$跳邻居拓扑和属性$r \geq 2$的函数。图神经网络和集体推理方法(例如,置信传播)依赖于来自直到$r$跳远的信息。在这项工作中,我们研究,如果使用更强大的置换不变的功能,有时可以避免需要分类收集信息以外的$1$跳。为此,我们引入了一种新的架构,即Set Twister,它概括了DeepSets(Zaheer等人,2017年),一个简单而广泛使用的置换不变表示。Set Twister理论上增加了DeepSets的表达能力,使其能够捕获高阶依赖关系,同时保持其简单性和低计算成本。从经验上讲,我们看到Set Twister在DeepSets上的准确性提高,以及各种图神经网络和集体推理方案在几个任务中的准确性提高,同时展示了其实现的简单性和计算效率。

英文摘要:

Node classification is a central task in relational learning, with the current state-of-the-art hinging on two key principles: (i) predictions are permutation-invariant to the ordering of a node's neighbors, and (ii) predictions are a function of the node's $r$-hop neighborhood topology and attributes, $r \geq 2$. Both graph neural networks and collective inference methods (e.g., belief propagation) rely on information from up to $r$-hops away. In this work, we study if the use of more powerful permutation-invariant functions can sometimes avoid the need for classifiers to collect information beyond $1$-hop. Towards this, we introduce a new architecture, the Set Twister, which generalizes DeepSets (Zaheer et al., 2017), a simple and widely-used permutation-invariant representation. Set Twister theoretically increases expressiveness of DeepSets, allowing it to capture higher-order dependencies, while keeping its simplicity and low computational cost. Empirically, we see accuracy improvements of Set Twister over DeepSets as well as a variety of graph neural networks and collective inference schemes in several tasks, while showcasing its implementation simplicity and computational efficiency.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑