Classical graph algorithms work well for combinatorial problems that can be thoroughly formalized and abstracted. Once the algorithm is derived, it generalizes to instances of any size. However, developing an algorithm that handles complex structures and interactions in the real world can be challenging. Rather than specifying the algorithm, we can try to learn it from the graph-structured data. Graph Neural Networks (GNNs) are inherently capable of working on graph structures; however, they struggle to generalize well, and learning on larger instances is challenging. In order to scale, we focus on a recurrent architecture design that can learn simple graph problems end to end on smaller graphs and then extrapolate to larger instances. As our main contribution, we identify three essential techniques for recurrent GNNs to scale. By using (i) skip connections, (ii) state regularization, and (iii) edge convolutions, we can guide GNNs toward extrapolation. This allows us to train on small graphs and apply the same model to much larger graphs during inference. Moreover, we empirically validate the extrapolation capabilities of our GNNs on algorithmic datasets.
translated by 谷歌翻译
本文研究了辍学图神经网络(DAVERGNNS),一种旨在克服标准GNN框架的局限性的新方法。在DAMPGNNS中,我们在输入图上执行多个GNN运行,其中一些节点随机且独立地在这些运行中丢弃。然后,我们将这些运行的结果结合起来获得最终结果。我们证明DAMPGNN可以区分无法通过GNN的消息分隔的各种图形邻域。我们导出了确保可靠分布辍学所需的运行数量的理论界限,我们证明了有关DACKGNNS的表现能力和限制的若干特性。我们在实验上验证了我们对表现力的理论结果。此外,我们表明DOWNNNS在已建立的GNN基准上表现得很竞争。
translated by 谷歌翻译
图神经网络(GNN)已成功应用于许多真实世界静态图。但是,由于模型设计,评估设置和训练策略的局限性,静态图的成功尚未完全转化为动态图。具体而言,现有的动态GNN并不包含静态GNN的最新设计,从而限制了其性能。动态GNN的当前评估设置不能完全反映动态图的不断发展的性质。最后,用于动态GNN的常用训练方法是不可扩展的。在这里,我们提出了Roland,这是现实世界动态图的有效图表学习框架。 Roland框架的核心可以帮助研究人员轻松地将任何静态GNN重新用于动态图。我们的见解是将不同GNN层的节点嵌入视为分层节点状态,然后随着时间的推移将其反复更新。然后,我们为动态图引入了实时更高的评估设置,该设置模仿了现实世界中的用例,其中GNN正在做出预测并在滚动基础上进行更新。最后,我们通过增量训练和元学习提出了一种可扩展有效的训练方法,以动态GNN。我们在未来链接预测任务上对八个不同的动态图数据集进行了实验。在三个数据集的标准评估设置下,使用Roland框架建立的模型平均相对平均互惠等级(MRR)的平均相对平均值(MRR)改进。我们发现对较大数据集的最先进的基线经历了不可存储的错误,而Roland可以轻松地扩展到5600万个边缘的动态图。在使用ROLAND训练策略重新实现这些基准线后,Roland模型平均相对于基线相对相对改善了15.5%。
translated by 谷歌翻译
大多数图形神经网络(GNNS)无法区分某些图形或图中的某些节点。这使得无法解决某些分类任务。但是,在这些模型中添加其他节点功能可以解决此问题。我们介绍了几种这样的增强,包括(i)位置节点嵌入,(ii)规范节点ID和(iii)随机特征。这些扩展是由理论结果激励的,并通过对合成子图检测任务进行广泛测试来证实。我们发现位置嵌入在这些任务中的其他扩展大大超过了其他扩展。此外,位置嵌入具有更好的样品效率,在不同的图形分布上表现良好,甚至超过了地面真实节点位置。最后,我们表明,不同的增强功能在既定的GNN基准中都具有竞争力,并建议何时使用它们。
translated by 谷歌翻译
图形神经网络(GNN)已被密切应用于各种基于图的应用程序。尽管他们成功了,但手动设计行为良好的GNN需要巨大的人类专业知识。因此,发现潜在的最佳数据特异性GNN体系结构效率低下。本文提出了DFG-NAS,这是一种新的神经体系结构搜索(NAS)方法,可自动搜索非常深入且灵活的GNN体系结构。与大多数专注于微构造的方法不同,DFG-NAS突出了另一个设计级别:搜索有关原子传播的宏观构造(\ TextBf {\ Textbf {\ Texttt {p}}})和转换(\ texttt {\ textttt {\ texttt {\ texttt {\ texttt { T}})的操作被整合并组织到GNN中。为此,DFG-NAS为\ textbf {\ texttt {p-t}}}的排列和组合提出了一个新颖的搜索空间,该搜索空间是基于消息传播的散布,定义了四个自定义设计的宏观架构突变,并采用了进化性algorithm to to the Evolutionary algorithm进行有效的搜索。关于四个节点分类任务的实证研究表明,DFG-NAS优于最先进的手动设计和GNN的NAS方法。
translated by 谷歌翻译
我们提出了一个新的图形神经网络,我们称为AgentNet,该网络专为图形级任务而设计。 AgentNet的灵感来自子宫性算法,具有独立于图形大小的计算复杂性。代理Net的体系结构从根本上与已知图神经网络的体系结构不同。在AgentNet中,一些受过训练的\ textit {神经代理}智能地行走图,然后共同决定输出。我们提供了对AgentNet的广泛理论分析:我们表明,代理可以学会系统地探索其邻居,并且AgentNet可以区分某些甚至3-WL无法区分的结构。此外,AgentNet能够将任何两个图形分开,这些图在子图方面完全不同。我们通过在难以辨认的图和现实图形分类任务上进行合成实验来确认这些理论结果。在这两种情况下,我们不仅与标准GNN相比,而且与计算更昂贵的GNN扩展相比。
translated by 谷歌翻译
随着深度学习技术的快速发展,各种最近的工作试图应用图形神经网络(GNN)来解决诸如布尔满足(SAT)之类的NP硬问题,这表明了桥接机器学习与象征性差距的潜力。然而,GNN预测的解决方案的质量并未在文献中进行很好地研究。在本文中,我们研究了GNNS在学习中解决最大可满足性(MaxSAT)问题的能力,从理论和实践角度来看。我们构建了两种GNN模型来学习来自基准的MaxSAT实例的解决方案,并显示GNN通过实验评估解决MaxSAT问题的有吸引力。我们还基于算法对准理论,我们还提出了GNNS可以在一定程度上学会解决MaxSAT问题的影响的理论解释。
translated by 谷歌翻译
Recently, graph neural networks (GNNs) have revolutionized the field of graph representation learning through effectively learned node embeddings, and achieved state-of-the-art results in tasks such as node classification and link prediction. However, current GNN methods are inherently flat and do not learn hierarchical representations of graphs-a limitation that is especially problematic for the task of graph classification, where the goal is to predict the label associated with an entire graph. Here we propose DIFFPOOL, a differentiable graph pooling module that can generate hierarchical representations of graphs and can be combined with various graph neural network architectures in an end-to-end fashion. DIFFPOOL learns a differentiable soft cluster assignment for nodes at each layer of a deep GNN, mapping nodes to a set of clusters, which then form the coarsened input for the next GNN layer. Our experimental results show that combining existing GNN methods with DIFFPOOL yields an average improvement of 5-10% accuracy on graph classification benchmarks, compared to all existing pooling approaches, achieving a new state-of-the-art on four out of five benchmark data sets.
translated by 谷歌翻译
图形神经网络(GNNS)已成为旨在对图形结构数据进行学习和推断的引人注目的模型,但是在理解GNN的基本限制方面几乎没有做出的工作,该限制可扩展到较大的图形并推广到分布外输入。 。在本文中,我们使用一个随机图生成器,该生成器使我们能够系统地研究图形大小和结构属性如何影响GNN的预测性能。我们提供的具体证据表明,在许多图形属性中,节点度分布的平均值和模态是确定GNN是否可以推广到看不见的图的关键特征。因此,我们使用多个节点更新功能和内部循环优化作为对汇总输入的单一类型的规范非线性转换的概括,提出了灵活的GNN(flex-gnn),并将内部循环优化作为概括,从而使网络可以灵活地适应新图。 Flex-GNN框架改善了几个推理任务的培训设置的概括。
translated by 谷歌翻译
Neural message passing algorithms for semi-supervised classification on graphs have recently achieved great success. However, for classifying a node these methods only consider nodes that are a few propagation steps away and the size of this utilized neighborhood is hard to extend. In this paper, we use the relationship between graph convolutional networks (GCN) and PageRank to derive an improved propagation scheme based on personalized PageRank. We utilize this propagation procedure to construct a simple model, personalized propagation of neural predictions (PPNP), and its fast approximation, APPNP. Our model's training time is on par or faster and its number of parameters on par or lower than previous models. It leverages a large, adjustable neighborhood for classification and can be easily combined with any neural network. We show that this model outperforms several recently proposed methods for semi-supervised classification in the most thorough study done so far for GCN-like models. Our implementation is available online. 1
translated by 谷歌翻译
组合优化是运营研究和计算机科学领域的一个公认领域。直到最近,它的方法一直集中在孤立地解决问题实例,而忽略了它们通常源于实践中的相关数据分布。但是,近年来,人们对使用机器学习,尤其是图形神经网络(GNN)的兴趣激增,作为组合任务的关键构件,直接作为求解器或通过增强确切的求解器。GNN的电感偏差有效地编码了组合和关系输入,因为它们对排列和对输入稀疏性的意识的不变性。本文介绍了对这个新兴领域的最新主要进步的概念回顾,旨在优化和机器学习研究人员。
translated by 谷歌翻译
用于图形组合优化问题的神经网络溶剂的端到端培训,例如旅行销售人员问题(TSP)最近看到了感兴趣的激增,但在几百节节点的图表中保持棘手和效率低下。虽然最先进的学习驱动的方法对于TSP在培训的古典索引时与古典求解器密切相关,但它们无法通过实际尺度的实际情况概括到更大的情况。这项工作提出了一个端到端的神经组合优化流水线,统一几个卷纸,以确定促进比在训练中看到的实例的概括的归纳偏差,模型架构和学习算法。我们的受控实验提供了第一个原则上调查这种零拍摄的概括,揭示了超越训练数据的推断需要重新思考从网络层和学习范例到评估协议的神经组合优化流水线。此外,我们分析了深入学习的最近进步,通过管道的镜头路由问题,并提供新的方向,以刺激未来的研究。
translated by 谷歌翻译
增强图在正规化图形神经网络(GNNS)方面起着至关重要的作用,该图形以信息传递的形式利用沿图的边缘进行信息交换。由于其有效性,简单的边缘和节点操作(例如,添加和删除)已被广泛用于图表增强中。然而,这种常见的增强技术可以显着改变原始图的语义,从而导致过度侵略性增强,从而在GNN学习中拟合不足。为了解决掉落或添加图形边缘和节点引起的此问题,我们提出了SoftEdge,将随机权重分配给给定图的一部分以进行增强。 SoftEdge生成的合成图保持与原始图相同的节点及其连接性,从而减轻原始图的语义变化。我们从经验上表明,这种简单的方法获得了与流行节点和边缘操纵方法的卓越精度,并且具有明显的弹性,可抵御GNN深度的准确性降解。
translated by 谷歌翻译
许多现代神经架构的核心的卷积运算符可以有效地被视为在输入矩阵和滤波器之间执行点产品。虽然这很容易适用于诸如图像的数据,其可以在欧几里德空间中表示为常规网格,延伸卷积操作者以在图形上工作,而是由于它们的不规则结构而被证明更具有挑战性。在本文中,我们建议使用图形内部产品的图形内核,即在图形上计算内部产品,以将标准卷积运算符扩展到图形域。这使我们能够定义不需要计算输入图的嵌入的完全结构模型。我们的架构允许插入任何类型和数量的图形内核,并具有在培训过程中学到的结构面具方面提供一些可解释性的额外益处,类似于传统卷积神经网络中的卷积掩模发生的事情。我们执行广泛的消融研究,调查模型超参数的影响,我们表明我们的模型在标准图形分类数据集中实现了竞争性能。
translated by 谷歌翻译
图形神经网络(GNNS)依赖于图形结构来定义聚合策略,其中每个节点通过与邻居的信息组合来更新其表示。已知GNN的限制是,随着层数的增加,信息被平滑,压扁并且节点嵌入式变得无法区分,对性能产生负面影响。因此,实用的GNN模型雇用了几层,只能在每个节点周围的有限邻域利用图形结构。不可避免地,实际的GNN不会根据图的全局结构捕获信息。虽然有几种研究GNNS的局限性和表达性,但是关于图形结构数据的实际应用的问题需要全局结构知识,仍然没有答案。在这项工作中,我们通过向几个GNN模型提供全球信息并观察其对下游性能的影响来认证解决这个问题。我们的研究结果表明,全球信息实际上可以为共同的图形相关任务提供显着的好处。我们进一步确定了一项新的正规化策略,导致所有考虑的任务的平均准确性提高超过5%。
translated by 谷歌翻译
A prominent paradigm for graph neural networks is based on the message passing framework. In this framework, information communication is realized only between neighboring nodes. The challenge of approaches that use this paradigm is to ensure efficient and accurate \textit{long distance communication} between nodes, as deep convolutional networks are prone to over-smoothing. In this paper, we present a novel method based on time derivative graph diffusion (TIDE), with a learnable time parameter. Our approach allows to adapt the spatial extent of diffusion across different tasks and network channels, thus enabling medium and long-distance communication efficiently. Furthermore, we show that our architecture directly enables local message passing and thus inherits from the expressive power of local message passing approaches. We show that on widely used graph benchmarks we achieve comparable performance and on a synthetic mesh dataset we outperform state-of-the-art methods like GCN or GRAND by a significant margin.
translated by 谷歌翻译
Pre-publication draft of a book to be published byMorgan & Claypool publishers. Unedited version released with permission. All relevant copyrights held by the author and publisher extend to this pre-publication draft.
translated by 谷歌翻译
图形神经网络(GNNS)最流行的设计范例是1跳消息传递 - 反复反复从1跳邻居聚集特征。但是,1-HOP消息传递的表达能力受Weisfeiler-Lehman(1-WL)测试的界定。最近,研究人员通过同时从节点的K-Hop邻居汇总信息传递到K-HOP消息。但是,尚无分析K-Hop消息传递的表达能力的工作。在这项工作中,我们从理论上表征了K-Hop消息传递的表达力。具体而言,我们首先正式区分了两种k-hop消息传递的内核,它们在以前的作品中经常被滥用。然后,我们通过表明它比1-Hop消息传递更强大,从而表征了K-Hop消息传递的表现力。尽管具有较高的表达能力,但我们表明K-Hop消息传递仍然无法区分一些简单的常规图。为了进一步增强其表现力,我们引入了KP-GNN框架,该框架通过利用每个跳跃中的外围子图信息来改善K-HOP消息。我们证明,KP-GNN可以区分几乎所有常规图,包括一些距离常规图,这些图无法通过以前的距离编码方法来区分。实验结果验证了KP-GNN的表达能力和有效性。 KP-GNN在所有基准数据集中都取得了竞争成果。
translated by 谷歌翻译
Deep learning has revolutionized many machine learning tasks in recent years, ranging from image classification and video processing to speech recognition and natural language understanding. The data in these tasks are typically represented in the Euclidean space. However, there is an increasing number of applications where data are generated from non-Euclidean domains and are represented as graphs with complex relationships and interdependency between objects. The complexity of graph data has imposed significant challenges on existing machine learning algorithms. Recently, many studies on extending deep learning approaches for graph data have emerged. In this survey, we provide a comprehensive overview of graph neural networks (GNNs) in data mining and machine learning fields. We propose a new taxonomy to divide the state-of-the-art graph neural networks into four categories, namely recurrent graph neural networks, convolutional graph neural networks, graph autoencoders, and spatial-temporal graph neural networks. We further discuss the applications of graph neural networks across various domains and summarize the open source codes, benchmark data sets, and model evaluation of graph neural networks. Finally, we propose potential research directions in this rapidly growing field.
translated by 谷歌翻译
Graph classification is an important area in both modern research and industry. Multiple applications, especially in chemistry and novel drug discovery, encourage rapid development of machine learning models in this area. To keep up with the pace of new research, proper experimental design, fair evaluation, and independent benchmarks are essential. Design of strong baselines is an indispensable element of such works. In this thesis, we explore multiple approaches to graph classification. We focus on Graph Neural Networks (GNNs), which emerged as a de facto standard deep learning technique for graph representation learning. Classical approaches, such as graph descriptors and molecular fingerprints, are also addressed. We design fair evaluation experimental protocol and choose proper datasets collection. This allows us to perform numerous experiments and rigorously analyze modern approaches. We arrive to many conclusions, which shed new light on performance and quality of novel algorithms. We investigate application of Jumping Knowledge GNN architecture to graph classification, which proves to be an efficient tool for improving base graph neural network architectures. Multiple improvements to baseline models are also proposed and experimentally verified, which constitutes an important contribution to the field of fair model comparison.
translated by 谷歌翻译