深度学习求深度图_关于图的深度学习成功挑战和下一步-CSDN博客

深度学习求深度图

重点 (Top highlight)

图神经网络的下一步是什么？ (What is next in store for graph neural networks?)

TL;DR This is the first in a series of posts where I will discuss the evolution and future trends in the field of deep learning on graphs.

TL; DR 这是 系列文章 中 的第一篇 ，我将讨论图上深度学习领域的演变和未来趋势。

Deep learning on graphs, also known as Geometric deep learning (GDL) [1], Graph representation learning (GRL), or relational inductive biases [2], has recently become one of the hottest topics in machine learning. While early works on graph learning go back at least a decade [3] if not two [4], it is undoubtedly the past few years’ progress that has taken these methods from a niche into the spotlight of the ML community and even to the popular science press (with Quanta Magazine running a series of excellent articles on geometric deep learning for the study of manifolds, drug discovery, and protein science).

d EEP学习曲线图上，也被称为几何深度学习(GDL)[1]，图表示学习(GRL)，或关系感应偏压[2]，最近已成为机器学习最热门的话题之一。虽然关于图学习的早期工作至少可以追溯到十年[3]甚至不是两个[4]，但无疑是过去几年的进步，使这些方法从利基市场逐渐成为ML社区甚至是ML社区的关注焦点。大众科学出版社(《 Quanta杂志》刊登了一系列有关几何深度学习的优秀文章，用于研究流形，药物发现和蛋白质科学 )。

Graphs are powerful mathematical abstractions that can describe complex systems of relations and interactions in fields ranging from biology and high-energy physics to social science and economics. Since the amount of graph-structured data produced in some of these fields nowadays is enormous (prominent examples being social networks like Twitter and Facebook), it is very tempting to try to apply deep learning techniques that have been remarkably successful in other data-rich settings.

图形是强大的数学抽象，可以描述从生物学和高能物理学到社会科学和经济学等领域的复杂的关系和相互作用系统。由于当今在这些领域中的某些领域产生的图结构化数据量巨大(突出的示例是诸如Twitter和Facebook之类的社交网络)，因此尝试将深度学习技术应用到其他数据丰富的领域中非常成功的尝试非常具有诱惑力。设置。

There are multiple flavours to graph learning problems that are largely application-dependent. One dichotomy is between node-wise and graph-wise problems, where in the former one tries to predict properties of individual nodes in the graph (e.g. identify malicious users in a social network), while in the latter one tries to make a prediction about the entire graph (e.g. predict solubility of a molecule). Furthermore, like in traditional ML problems, we can distinguish between supervised and unsupervised (or self-supervised) settings, as well as transductive and inductive problems.

可以用多种方式来描述学习问题，这些问题很大程度上取决于应用程序。一种二分法介于节点问题和图问题之间，其中前者试图预测图中单个节点的属性(例如，识别社交网络中的恶意用户)，而后者则试图对以下问题做出预测：整个图形(例如，预测分子的溶解度)。此外，像传统的机器学习问题一样，我们可以区分有监督和无监督 (或自我监督 )设置，以及传导性和归纳性问题。

Similarly to convolutional neural networks used in image analysis and computer vision, the key to efficient learning on graphs is designing local operations with shared weights that do message passing [5] between every node and its neighbours. A major difference compared to classical deep neural networks dealing with grid-structured data is that on graphs such operations are permutation-invariant, i.e. independent of the order of neighbour nodes, as there is usually no canonical way of ordering them.

与图像分析和计算机视觉中使用的卷积神经网络相似，在图上进行高效学习的关键是设计具有共享权重的本地操作，该权重在每个节点及其邻居之间进行消息传递[5]。与处理网格结构数据的经典深度神经网络相比，主要区别在于在图上这种操作是 不变排列 ，即与邻居节点的顺序无关，因为通常没有规范的排序方式。

Despite their promise and a series of success stories of graph representation learning (among which I can selfishly list the acquisition by Twitter of the graph-based fake news detection startup Fabula AI I have founded together with my students), we have not witnessed so far anything close to the smashing success convolutional networks have had in computer vision. In the following, I will try to outline my views on the possible reasons and how the field could progress in the next few years.

尽管他们的承诺和一系列图形表示学习的成功案例(其中我可以自私地列出Twitter与我和学生们一起创建的基于图形的假新闻检测初创公司Fabula AI的收购 )，但到目前为止我们还没有目睹过卷积网络在计算机视觉方面所取得的成功均非凡。在下文中，我将尝试就可能的原因以及该领域在未来几年中的进展情况概述我的观点。

Standardised benchmarks like ImageNet were surely one of the key success factors of deep learning in computer vision, with some [6] even arguing that data was more important than algorithms for the deep learning revolution. We have nothing similar to ImageNet in scale and complexity in the graph learning community yet. The Open Graph Benchmark launched in 2019 is perhaps the first attempt toward this goal trying to introduce challenging graph learning tasks on interesting real-world graph-structured datasets. One of the hurdles is that tech companies producing diverse and rich graphs from their users’ activity are reluctant to share these data due to concerns over privacy laws such as GDPR. A notable exception is Twitter that made a dataset of 160 million tweets with corresponding user engagement graphs available to the research community under certain privacy-preserving restrictions as part of the RecSys Challenge. I hope that many companies will follow suit in the future.

像ImageNet这样的标准化基准无疑是计算机视觉深度学习成功的关键因素之一，甚至有人[6]认为，数据比深度学习革命的算法更重要。在图学习社区中，我们在规模和复杂性上都没有与ImageNet类似的东西。 2019年启动的Open Graph Benchmark可能是实现此目标的首次尝试，试图在有趣的现实世界中以图结构化的数据集引入具有挑战性的图学习任务。障碍之一是，由于对隐私法(例如GDPR)的担忧，高科技公司不愿从用户的活动中生成多样而丰富的图表。作为一个著名的例外，Twitter作为RecSys Challenge的一部分，在一定的隐私保护限制下，向研究社区提供了1.6亿条推文的数据集，以及相应的用户参与图。我希望将来有许多公司效仿。

Software libraries available in the public domain played a paramount role in “democratising” deep learning and making it a popular tool. If until recently, graph learning implementations were primarily a collection of poorly written and scarcely tested code, nowadays there are libraries such as PyTorch Geometric or Deep Graph Library (DGL) that are professionally written and maintained with the help of industry sponsorship. It is not uncommon to see an implementation of a new graph deep learning architecture weeks after it appears on arxiv.

公共领域中可用的软件库在“民主化”深度学习并使之成为一种流行的工具中发挥了至关重要的作用。如果直到最近，图形学习的实现方式主要是编写不良且很少经过测试的代码的集合，那么如今，诸如PyTorch Geometric或Deep Graph Library(DGL)之类的库是在行业赞助的帮助下进行专业编写和维护的。在arxiv上出现新的图深度学习架构几周后，这种情况的实现并不少见。

Scalability is one of the key factors limiting industrial applications that often need to deal with very large graphs (think of Twitter social network with hundreds of millions of nodes and billions of edges) and low latency constraints. The academic research community has until recently almost ignored this aspect, with many models described in the literature completely inadequate for large-scale settings. Furthermore, graphics hardware (GPU), whose happy marriage with classical deep learning architectures was one of the primary forces driving their mutual success, is not necessarily the best fit for graph-structured data. In the long run, we might need specialised hardware for graphs [7].

可伸缩性是限制工业应用程序的关键因素之一，工业应用程序通常需要处理非常大的图形(想想具有数亿个节点和数十亿个边缘的Twitter社交网络)和低延迟约束。直到最近，学术研究界几乎忽略了这一方面，文献中描述的许多模型完全不足以用于大规模环境。此外，图形硬件(GPU)与经典的深度学习体系结构的幸福婚姻是驱动其相互成功的主要力量之一，但它不一定最适合图结构化数据。从长远来看，我们可能需要专用的图形硬件[7]。

Dynamic graphs are another aspect that is scarcely addressed in the literature. While graphs are a common way of modelling complex systems, such an abstraction is often too simplistic as real-world systems are dynamic and evolve in time. Sometimes it is the temporal behaviour that provides crucial insights about the system. Despite some recent progress, designing graph neural network models capable of efficiently dealing with continuous-time graphs represented as a stream of node- or edge-wise events is still an open research question.

动态图是文献中很少涉及的另一方面。虽然图是对复杂系统建模的一种常见方式，但是由于现实世界中的系统是动态的并且随时间变化的，所以这种抽象通常过于简单。有时正是时间行为提供了有关系统的关键见解。尽管最近取得了一些进展，但设计能够有效处理以节点或边沿事件流表示的连续时间图的图神经网络模型仍然是一个尚待研究的问题。

Higher-order structures such as motifs, graphlets, or simplicial complexes are known to be of importance in complex networks, e.g. describing protein-protein interactions in biological applications. Yet, the majority of graph neural networks are limited to nodes and edges only. Incorporating such structures into the message passing mechanism could bring more expressive power to graph-based models.

已知诸如图案，小图或简单复合物之类的高级结构在复杂网络中很重要，例如描述生物学应用中的蛋白质-蛋白质相互作用。然而，大多数图神经网络只限于节点和边。将这样的结构合并到消息传递机制中可以为基于图形的模型带来更多的表达能力。

Theoretical understanding of the expressivity of graph neural networks is rather limited. It is common to see both dramatic boosts in performance coming from the use of graph neural networks in some settings along with almost no difference in others. It is not completely clear yet when and why graph neural networks work well or fail. The problem is difficult because one has to consider both the structure of the underlying graph as well as the data on it. For graph classification problems concerned with the graph connectivity only, recent works showed that graph neural networks are equivalent to the Weisfeiler-Lehman graph isomorphism test [8] (a heuristic for solving a classical problem in graph theory of determining if two graphs are identical up to the permutation of their nodes). This formalism sheds light on why, for example, graph neural networks fail on instances of non-isomorphic graphs that cannot be distinguished by this simple test. Going beyond the Weisfeiler-Lehman hierarchy of tests while keeping the low linear complexity that makes graph neural networks so attractive is an open research question.

对图神经网络表达能力的理论理解相当有限。常见的是，在某些情况下使用图神经网络既可以显着提高性能，又可以在其他情况下几乎没有区别。尚不清楚何时何地图神经网络运行良好或失败。这个问题很困难，因为必须同时考虑基础图形的结构以及其中的数据。对于仅与图连通性有关的图分类问题，最近的工作表明，图神经网络等效于Weisfeiler-Lehman图同构检验 [8](一种用于解决图论中确定两个图是否相同的经典问题的启发式方法)到其节点的排列)。这种形式主义阐明了为什么图神经网络在非同构图实例上失败的原因，这些实例无法通过此简单测试来区分。超越测试的Weisfeiler-Lehman层次结构，同时保持低线性复杂度(使图神经网络如此具有吸引力)是一个开放的研究问题。

Robustness and guaranteed performance of graph neural networks in the presence of noisy data or when subjected to adversarial attacks [9] is another interesting and largely virgin research field.

图神经网络在存在嘈杂数据或受到对抗攻击时的鲁棒性和有保证的性能 [9]是另一个有趣且原始的研究领域。

Applications are perhaps the most gratifying part of the field. Having worked on graph learning for many years now, I have made friends with particle physicists [10], clinicians [11], biologists and chemists [12] — people I would unlikely meet otherwise had we not worked on applications in their respective fields. If I were to bet on just one field where graph deep learning could make the biggest impact in the next few years, I would indicate structural biology and chemistry. In these fields, graph-based models can be used both as low-level models for molecules [5], as well as high-level models of interactions between them [13,11]. Combining these might be the key to getting to a level that would be useful for the pharmaceutical industry — and we see initial signs of this, with graph neural networks used to discover a new class of antibiotics [14] earlier this year or to predict interactions between proteins [12]. If graph deep learning delivers on its promise, the traditionally very long and astoundingly expensive process of discovering, developing, and testing new drugs may never look the same.

应用程序可能是该领域中最令人满意的部分。从事图形学习已经很多年了，我已经与粒子物理学家[10]，临床医生[11]，生物学家和化学家[12]成为朋友–如果我们不从事各自领域的应用程序工作，我将不太可能见面。如果我只赌图深度学习可能在未来几年产生最大影响的领域，我将指出结构生物学和化学领域。在这些领域中，基于图的模型既可以用作分子的低级模型[5]，也可以用作它们之间相互作用的高级模型[13,11]。将这些结合起来可能是达到对制药业有用的水平的关键-我们看到了这一初步迹象，利用图神经网络在今年早些时候发现了一类新的抗生素[14]或预测相互作用蛋白质之间[12]。如果图式深度学习兑现了其诺言，那么传统上非常漫长且惊人的昂贵发现，开发和测试新药的过程可能看起来就不一样了。

[1] M. M. Bronstein et al. Geometric deep learning: going beyond Euclidean data (2017), IEEE Signal Processing Magazine 34(4):18–42.

[1] MM Bronstein等。几何深度学习：超越欧几里得数据 (2017年)，IEEE信号处理杂志34(4)：18–42。

[2] P. Battaglia et al., Relational inductive biases, deep learning, and graph networks (2018), arXiv:1806.01261.

[2] P.Battaglia等人，《关系归纳偏差，深度学习和图网络》 (2018年)，arXiv：1806.01261。

[3] F. Scarselli et al. The graph neural network model (2008), IEEE Transactions on Neural Networks 20(1):61–80.

[3] F. Scarselli等。图神经网络模型(2008)，IEEE Transactions on Neural Networks 20(1)：61–80。

[4] A. Küchler, C. Goller (1996). Inductive learning in symbolic domains using structure-driven recurrent neural networks (1996), Proc. Künstliche Intelligenz.

[4] A.Küchler，C. Goller(1996)。使用结构驱动的递归神经网络(1996)，Proc。 KünstlicheIntelligenz。

[5] J. Gilmer et al., Neural message passing for quantum chemistry (2017), ICML.

[5] J. Gilmer等人，《量子化学的神经信息传递》 (2017年)，ICML。

[6] A. Wissner-Gross, Datasets over algorithms (2016).

[6] A. Wissner-Gross，“ 算法上的数据集” (2016年)。

[7] C.-Y. Gui et al.. A survey on Graph Processing accelerators: Challenges and Opportunities (2019), arXiv:1902.10130.

[7] C.-Y. Gui等人。关于图处理加速器的调查：挑战与机遇 (2019)，arXiv：1902.10130。

[8] K. Xu et al. How powerful are graph neural networks? (2019), ICLR.

[8] K. Xu等。图神经网络有多强大？ (2019)，ICLR。

[9] D. Zügner et al., Adversarial attacks on neural networks for graph data (2018), Proc. KDD.

[9]D.Zügner等人，针对图数据的神经网络的对抗攻击 (2018)，Proc。 KDD。

[10] N. Choma et al. Graph neural networks for IceCube signal classification (2018), Proc. ICMLA.

[10] N. Choma等。用于IceCube信号分类的图神经网络 (2018)，Proc。 ICMLA。

[11] K. Veselkov et al. HyperFoods: Machine intelligent mapping of cancer-beating molecules in foods (2019), Scientific Reports 9.

[11] K. Veselkov等。 HyperFoods：机器智能绘制食品中抗癌分子的图谱 (2019年)，《科学报告》 9。

[12] P. Gainza et al. Deciphering interaction fingerprints from protein molecular surfaces using geometric deep learning (2020), Nature Methods 17:184–192.

[12] P. Gainza等。使用几何深度学习 (2020)，自然方法17：184–192 解密蛋白质分子表面的相互作用指纹。

[13] M. Zitnik et al. Modeling polypharmacy side effects with graph convolutional networks (2018), Bioinformatics 34(13):457–466.

[13] M. Zitnik等。使用图卷积网络建模多药副作用 (2018年)，生物信息学34(13)：457–466。

[14] J. Stokes et al. A deep learning approach to antibiotic discovery (2020), Cell, 180(4).

[14] J. Stokes等。抗生素发现的深度学习方法(2020)，Cell，180(4)。

A Chinese translation of this post is available by courtesy of Zhiyong Liu. Interested in Graph Deep Learning? See my other posts on Medium, or follow me on Twitter.

一个 中国的翻译 这一职位的是通过提供礼貌 志勇刘 。 对图深度学习感兴趣吗？ 在Medium上 查看我的 其他帖子 ，或在 Twitter上 关注我 。