ICML2020 《When Does Self-Supervision Help Graph Convolutional Networks?》 Reading Notes

Paper Introduction 本文是自监督方法在GCNs上首次系统的探索,设计了3种自监督任务来将分析自监督在GCN中起到的作用。自监督旨在充分利用unlabeled数据中的知识来设计前置任务(pretext task),来帮助模型学习更具迁移性和泛化能力的表示。前置任务可以认为是对目标任务有帮助的辅助正则化网络,设计用于帮助原任务学习到更多下游任务相关的语义信息。 GCN任务通常是直推半监督的(transductive semi-supervised),含有大量unlabeled数据,而self-supervision(SSL)可以充分利用unlabeled data, 那么就产生了一个值得探索的问题:将自监督学习应用到GCN上是否也可以达到提升泛化能力和鲁棒能力的效果? 先给结论 Q1: 自监督学习可否在分类任务中提升GCN? 如果可以,如何将其合并到 GCN 中以最大化增益? A1: 本文证明了通过多任务学习将自监督学习融入 GCN 是有效的,即多任务损失作为 GCN 训练中的正则化项。 这种作为自监督作为正则化项的方法,强于用自监督来预训练或者self-training。 Q2: 前置任务的设计重要吗? GCN 有哪些有用的自监督前置任务? A2: 本文研究了三个基于图属性的自监督任务。 分别是节点聚类node clustering, 图划分graph partitioning 和图补全graph completion。 并且进一步说明不同的模型和数据集倾向于不同的自监督任务。 Q3: 自监督也会影响 GCN 的对抗鲁棒性吗? 如果是,如何设计前置任务? A3: 本文进一步将上述发现推广到对抗性训练环境中。提供了广泛的结果,以表明自监督还可以提高 GCN 在各种攻击下的鲁棒性,而不需要更大的模型或额外的数据。 Method GCNs $\boldsymbol{Z}=\hat{\boldsymbol{A}} \operatorname{ReLU}\left(\hat{\boldsymbol{A}} \boldsymbol{X} \boldsymbol{W}_{0}\right) \boldsymbol{W}_{1}$可以分为两块来看 (1) 特征提取模块$f_{\theta}(\boldsymbol{X}, \hat{\boldsymbol{A}}) = \hat{\boldsymbol{A}} \operatorname{ReLU}\left(\hat{\boldsymbol{A}} \boldsymbol{X} \boldsymbol{W}_{0}\right)$ 参数为$\theta = \{\boldsymbol{W}_{0}\}$和(2)线性变换模块$\boldsymbol{Z}=f_{\theta}(\boldsymbol{X}, \hat{\boldsymbol{A}}) \boldsymbol{\Theta}$ 其中 参数$ \boldsymbol{\Theta} = \boldsymbol{W}_{1}$。 半监督GCN优化任务的目标函数为: $$ \begin{aligned} \boldsymbol{Z} &=f_{\theta}(\boldsymbol{X}, \hat{\boldsymbol{A}}) \boldsymbol{\Theta} \\ \theta^{*}, \boldsymbol{\Theta}^{} &=\arg \min_{\theta, \boldsymbol{\Theta}} \mathcal{L}_{\mathrm{sup}}(\theta, \boldsymbol{\Theta}) \\ &=\arg \min_{\theta, \boldsymbol{\Theta}} \frac{1}{\left|\mathcal{V}_{\text {label }}\right|} \sum_{v_{n} \in \mathcal{V}_{\text {label }}} L\left(\boldsymbol{z}_{n}, \boldsymbol{y}_{n}\right) \end{aligned} \tag{1} $$ 其中$L(\cdot, \cdot)$是每个labeled node的损失函数。 ...

April 8, 2022 · 3 min · JhuoW

Proximal Gradient Descent

当目标函数中有不可微部分时,可使用近端梯度下降来优化(Proximal Gradient Descent) 假设目标函数如下: $$ \definecolor{energy}{RGB}{114,0,172} \definecolor{freq}{RGB}{45,177,93} \definecolor{spin}{RGB}{251,0,29} \definecolor{signal}{RGB}{18,110,213} \definecolor{circle}{RGB}{217,86,16} \definecolor{average}{RGB}{203,23,206} \definecolor{red}{RGB}{255,0,0} f(w) = g(w) + h(w) $$ 其中$g(w)$是可微凸函数,$h(w)$是不可微(或局部不可微)凸函数。 以线性回归为例, 给定$X \in \mathbb{R}^{m \times n}$, $y \in \mathbb{R}^m$, Ridge Regression的目标函数为 $$ f(\boldsymbol{w})=\underbrace{\frac{1}{2}||\boldsymbol{y}-\boldsymbol{X} \boldsymbol{w}||_{2}^{2}}_{g(\boldsymbol{w})}+\underbrace{\lambda||\boldsymbol{w}||_{2}}_{h(\boldsymbol{w})} $$ 因为$\ell_2$ norm处处可导,所以Ridge可以用SGD或GD来直接优化。但是若目标函数为Lasso,即正则化项定义为$\ell_1$ norm: $$ f(\boldsymbol{w})=\underbrace{\frac{1}{2}||\boldsymbol{y}-\boldsymbol{X} \boldsymbol{w}||_{2}^{2}}_{g(\boldsymbol{w})}+\underbrace{\lambda||\boldsymbol{w}||_{1}}_{h(\boldsymbol{w})} $$ 这里$h(w)=\lambda||\boldsymbol{w}||_{1}$在$w=0$处不可导,那么可用PGD来优化。 Proximity Operator 近端算子: 对于不可微函数$h(w)$, $h(w)$的proximity operator定义为: $$ u^* = \operatorname{prox}_{\color{signal}h}(w)=\underset{u}{\arg \min }\left(h(u)+\frac{1}{2}||u-w||_{2}^{2}\right) $$ 近端算子$\operatorname{prox}_{\color{signal}h}(w)$只和不可微凸函数$h(\cdot)$有关。 上式含义,给定一个不可微凸函数$h(\cdot)$, 给定向量$w \in \mathbb{R}^n$, 找到向量$u = u^*$, 使得公式$h(u)+\frac{1}{2}||u-w||_{2}^{2}$最小。 这个$u^* = \operatorname{prox}_{\color{signal}h}(w)$就是$h(\cdot)$在给定$w$条件下的近端算子(Proximity Operator)。$u^* = \operatorname{prox}_{\color{signal}h}(w)$要求最佳的$u^* $可以使得函数值$h(u^*)$尽可能小,同时$u^*$要尽可能接近给定的$w$。 ...

April 4, 2022 · 3 min · JhuoW

Everything about Graph Laplacian

Introduction The spectral graph theory studies the properties of graphs via the eigenvalues and eigenvectors of their associated graph matrices: the adjacency matrix and the graph Laplacian and its variants. The Laplacian allows a natural link between discrete representations, such as graphs, and continuous representations, such as vector spaces and manifolds. The most important application of the Laplacian is spectral clustering that corresponds to a computationally tractable solution to the graph partitionning problem. Another application is spectral matching that solves for graph matching. ...

April 2, 2022 · 16 min · JhuoW

NeurIPS2021 《Topology-Imbalance Learning for Semi-Supervised Node Classification》 Reading Notes

Paper Introduction 类别不均衡(Class Imbalance)是真实场景中非常常见的问题。一般在我们提及类别不均衡时,默认指的是数量不均衡:即不同类中训练样本数量的不一致带来的模型于不同类别学习能力的差异,由此引起的一个严重问题是模型的决策边界会主要由数量多的类来决定 。 但是在图结构中,不同类别的训练样本不仅有在数量上的差异,也有在位置结构上的差异.这就使得图上的类别不均衡问题有了一个独特的来源:拓扑不均衡。这个工作最主要的动机就是研究拓扑不均衡的特点,危害以及解决方法,希望能够引起社区对拓扑不均衡问题的重视。 本文提出Topology-Imbalance Node Representation Learning (TINL), 主要关注拓扑不平衡导致的决策边界漂移。所谓拓扑不平衡值得是, labeled nodes的位置如果位于拓扑中的决策边界,那么会传播错误的影响。 如上图所示,颜色和色调分别表示节点从labeled node接收到的influence类型和强度,节点R1位于两类节点的拓扑边界,第一张图可以看出,两个$\mathbf{x}$节点面临influence conflict问题,两个$\mathbf{Y}$节点由于远离R2,面临影响力不足的问题。也就是,如果决策便捷有labeled node(如R1), 那么他的影响力很容易传播给另一个类的边界unlabeled节点,导致影响力冲突,从而分类错误。 而冲突较小的labeled node更可能位于类的拓扑中心(如R2),顾增加其权重,是的它在训练过程中发挥更大作用。 Understanding Topology Imbalance via Label Propagation Label Propagation中,labels从labeled node延边传播, 看做label从labeled node开始的随机游走过程。LP最终收敛状态可以认为每个节点的soft-labels: $$ \boldsymbol{Y}=\alpha\left(\boldsymbol{I}-(1-\alpha) \boldsymbol{A}^{\prime}\right)^{-1} \boldsymbol{Y}^{0} $$ 其中$\boldsymbol{A}^{\prime}=\boldsymbol{D}^{-\frac{1}{2}} A D^{-\frac{1}{2}}$,其实就是PageRank的极限分布, $\boldsymbol{Y}^{0}$为每个节点的初始one-hot label。 第$i$个节点的预测结果为$\boldsymbol{q}_{i}=\arg \max _{j} \boldsymbol{Y}_{i j}$,每个节点的预测向量反映了每个节点主要受哪个类的影响。图(a)反映了GCN与LP的预测一致性,所以LP的节点影响力边界可以作为GNN的决策边界。理想状态下,labeled node的影响力边界应与真实类边界一致,例如红色的labeled node 在LP下所传播的影响力范围,应与所有红色node的范围一致。但是如图(b)所示,蓝色的labeled node如果较多位于真实类边界,这些位于边界的节点也会传播影响力,从而导致位于边界的真是红色节点被预测为蓝色,预测边界向红色类偏移。 Measuring Topology Imbalance by Influence Conflict 可以看出,位于决策边界的labeled node 会不可避免的将影响力传播到其他类节点,因此需要衡量labeled node与其所属类的相对拓扑位置(位于类边缘还是中心)。由于Homophily, 位于类边缘的节点也具有和其邻居相似的性质,因此利用邻域特征差别来判断labeled node是否位于边缘是不可靠的。因此本文利用整个图中的节点影响力冲突,提出基于冲突检测的拓扑相对位置Conflict Detection-based Topology Relative Location metric (Totoro). ...

April 2, 2022 · 2 min · JhuoW

Expectation Maximization

最大似然估计MLE 数据 $X = \{x_1, \cdots x_N\}$, 模型参数为$\theta$,Likelihood 定义为 $P(X | \theta)$:当参数为$\theta$时,观测到给定数据$X$的概率。 $$ P(X|\theta) = L(\theta | X) = P_\theta(X) \tag{1} $$ 最大似然估计 (Maximum Likelihood Estimation, MLE): $$ \theta_{\mathrm{MLE}} = \arg \max_\theta P(X|\theta) \tag{2} $$ 最大似然估计:给定一组样本$X$,模型的参数$\theta$是研究对象。若能找到参数$\theta_{\mathrm{MLE}}$,使得样本发生的可能性最大,则此估计值$\theta_{\mathrm{MLE}}$为参数$\theta$的最大似然估计。 举例来说,如果模型是单个Gaussian Distribution下,参数为Gaussian Distribution的参数(均值$\mu$, 标准差$\Sigma$, $\theta = {\mu, \Sigma}$). 给定一组数据$X$, 要计算$X$来自什么样的Gaussian,即:$P(\cdot | \theta) = f_\theta(\cdot) = \mathcal{N}(\cdot | \mu,\Sigma)$是一个Gaussian Distribution函数,目标为: $$ \theta_{\mathrm{MLE}} = \mu^\star, \Sigma^\star = \arg \max_{\mu,\Sigma} \sum^N_{i = 1} \log \mathcal{N}(x_i|\mu,\Sigma) \tag{3} $$ 即MLE的目标是找到最佳的高斯分布,是的从该分布中采样出数据$X$的概率最高。 ...

April 1, 2022 · 4 min · JhuoW

ICML2020 《Robust Graph Representation Learning via Neural Sparsification》 Reading Notes

Paper Introduction GNN的neighborhood aggregation中会引入邻域中任务无关的邻居,所以要移除图中有大量任务无关的边,顾本文提出NeuralSparse来解决该问题。 在构造数据集时,两个节点连接的动机可能与拿到这张图后要进行的下游任务无关,例如下游任务是节点分类,那么和这个任务相关的连接应是同类节点间产生边。但是构造图数据集是的两个节点连接的动机可能是特征相似的节点(不一定同类)。 下图给出一个例子,Blue 和 Red节点分别采样自两个独立的二维Gaussian distribution, 图1(a)显示两类节点的原始features有大量重合。对于每个节点,随机抽取10个其他节点作为他的邻居,这样生成的图(图1(b) )中的边和node label没有任何关系,即所有边都与节点分类任务无关。在这个图上做GCN得到(图1(b)下方)的node embedding,可以看到GCN学到的node embedding在任务无关的边上,无法区分两类节点。 图1(c)是DropEdge学到的node embedding,图1(d)为本文的NeuralSparse学到的node embedding。 Present work:根据监督信号来选择任务相关的边。 由sparsification network 和GNN两个模块组成. sparsification network旨在参数化稀疏过程,即在给定预算下, 为每个节点选择任务相关的一节邻居。在训练阶段,优化稀疏策略,即训练一个可以将图稀疏化的网络。在测试阶段,将测试图输入网络,得到一个稀疏化的图。 NeuralSparse Theoretical justification 首先从统计学习角度建模问题。将预测任务定义为$P(Y|G)$, 其中$Y$是预测目标,$G$是输入图。本文利用稀疏化子图来移除任务无关的边, 问题形式化为: $$ P(Y \mid G) \approx \sum_{g \in \mathbb{S}_{G}} P(Y \mid g) P(g \mid G) $$ $g$是一个稀疏化子图,$\mathbb{S}_{G}$为$G$的稀疏化子图集合。 $P(Y \mid g)$ 为给定一个稀疏化子图$g$,用$g$过GNN后预测为$Y$的概率。 $$ \sum_{g \in \mathbb{S}_{G}} P(Y \mid g) P(g \mid G) \approx \sum_{g \in \mathbb{S}_{G}} Q_{\theta}(Y \mid g) Q_{\phi}(g \mid G) $$ 用函数来近似(计算) 分布, 即给定$g$ 预测为$Y$的概率$ P(Y \mid g)$定义为一个参数为$\theta$的函数$Q_{\theta}(Y \mid g)$, 从$G$中获得子图$g$的概率$P(g \mid G)$定义为一个参数为$\phi$的函数$Q_{\phi}(g \mid G)$。 ...

April 1, 2022 · 2 min · JhuoW

NeurIPS2021 《Neo-GNNs:Neighborhood Overlap-aware Graph Neural Networks for Link Prediction》 Reading Notes

Paper Introduction 由于GNNs过度强调平滑的节点特征而不是图结构,使得在Link Prediction任务上的表现甚至弱于heuristic方法。平滑邻居难以反映邻域结构的相关性以及其他拓扑特征。 Structural information, (e.g., overlapped neighborhoods, degrees, and shortest path), is crucial for link prediction whereas GNNs heavily rely on smoothed node features rather than graph structure。 Link prediction heuristics: 基于预定义的假设的链路预测。举几个例子[1]: Common Neighbors (CN): 公共邻居较多的节点存在边(heuristic),则需要计算节点对间的公共邻居。 Preferential Attachment (PA): 一个节点当前的连接越多,那么它越有可能接受到新的连接(heuristic),这需要统计每个节点的度, i.e., $P A(x, y)=|N(x)| *|N(y)|$ Katz Index heuristic: $\sum^{\infty}_{\ell=1} \beta^{\ell}|walks(x,y)=\ell|$ 表示从$x$到$y$的所有路径数, $0<\beta<1$, 表示越长的路径权重越低。 katz Index作为Link prediction heuristic假设来作为边是否存在的预测。 本文提出了Neighborhood Overlap-aware Graph Neural Networks (Neo-GNNs)来从邻接矩阵中学习结构信息,并且估计重叠多跳邻域用于link prediction。 Preliminaries GNNs for Link Prediction $$ \hat{y}_{i j}=\sigma\left(s\left(h_{i}^{(L)}, h_{j}^{(L)}\right)\right) $$ ...

March 30, 2022 · 2 min · JhuoW

NeurIPS2021 《Representing Long-Range Context for Graph Neural Networks with Global Attention》 Reading Notes

Paper Introduction 加深GNN层数来增加感受野会导致优化不稳定性,比如梯度消失和oversmoothing。 因此本文采用Transformer-based self-attention来学习成对节点间的长距离关系,并且提出一种新型的readout池化机制来学习global graph embedding。即在一个GNN模块后接一个置换不变(permutation-invariant)Transformer, GNN模块捕获local信息,Transformer捕获global信息。 GNN作为一种专门的架构医学系节点直接邻域结构的局部表示, 而Transformer作为全局推理模块以位置无关的方式计算所有成对节点的交互。作者认为,一个没有positional encoding的Transformer是置换不变的,因此很适合图。 Motivation 强关系Inductive bias(我的理解是Homophily假设) 鼓励学习局部短距离的关联性。 而对于长距离相关性,结构化较低的模块(不需要过于考虑图的结构信息)更受欢迎。 GraphTrans leaves learning long-range dependencies to Transformer, 通过Transformer来学习图中所有节点对的依赖关系而不是只关注局部邻居。 下图中展示了一个子图的attention map。一共有17个节点,横坐标表示目标节点,纵坐标表示源节点,第$i$行第$j$列表示节点$i$在Transformer中聚合$j$的attention权重。第18行为一个特殊的$<CLS>$token 作为图的readout embedding。结合本文的SOTA效果,表面在学习长距离依赖时不考虑图结构先验(spatial priors)对Graph summarization(graph-level representation)是有必要的 Model GNN Module 一个通用的GNN模块: $$ \boldsymbol{h}_{v}^{\ell}=f_{\ell}\left(\boldsymbol{h}_{v}^{\ell-1},\left\{\boldsymbol{h}_{u}^{\ell-1} \mid u \in \mathcal{N}(v)\right\}\right), \quad \ell=1, \ldots, L_{\mathrm{GNN}} $$ Transformer Module 通过上面的GNN模块,我们可以得到每个节点的embedding $h_{v}^{L_{\mathrm{GNN}}}$, 将所有节点作为Transformer的Input。传统Transformer中输入先计算Self-attention(当前输入的$Q$向量和所有节点的$K$向量做内积得到输入节点和其他节点的att值,再用这个att值来为当前输入节点加权聚合所有节点的$V$向量),聚合后再和自身相加做residual,然后在做Layer Norm, 即对节点$i$的表示做Layer Norm 为$x_i^\prime = \frac{x_i-m}{\sigma}$, 其中$m$为$x_i$的均值, $\sigma$为$x_i$的标准差。 这里的Transformer不同的是, 先对所有节点做一次MLP,然后直接计算Layer Norm: $$ \overline{\boldsymbol{h}}_{v}^{0}=\operatorname{LayerNorm}\left(\boldsymbol{W}^{\text {Proj }} \boldsymbol{h}_{v}^{L_{\mathrm{GNN}}}\right) $$ 其中$\boldsymbol{W}^{\text {Proj }} \in \mathbb{R}^{d_{\mathrm{TF}} \times d_{L_{\mathrm{GNN}}}}$, 把GNN的输出维度转为TF的输入维度$d_{\mathrm{TF}}$。将所有节点的GNN node embeddings作为Transformer的输入(无positional encoding)。 每个节点的$Q$, $K$和$V$向量分别用$\boldsymbol{W}_{\ell}^{Q}, \boldsymbol{W}_{\ell}^{K}, \boldsymbol{W}_{\ell}^{V} \in \mathbb{R}^{d_{\mathrm{TF}} / n_{\text {head }} \times d_{\mathrm{TF}} / n_{\text {head }}}$计算, 对于第$\ell$层 Transformer, 节点$v$ 的$Q$向量$Q_v = \boldsymbol{W}_{\ell}^{Q} \overline{\boldsymbol{h}}_{v}^{\ell-1}$和节点$u$的$K$向量$K_u = \boldsymbol{W}_{\ell}^{K} \overline{\boldsymbol{h}}_{u}^{\ell-1}$做内积,得到两个节点之间的attention。然后用$\alpha_{v, u}^{\ell}$来为节点$v$聚合其他所有节点的$V$向量$V_u = \boldsymbol{W}_{\ell}^{V} \overline{\boldsymbol{h}}_{u}^{\ell-1}$, 如下所示: $$ a_{v, u}^{\ell}=\left(\boldsymbol{W}_{\ell}^{Q} \overline{\boldsymbol{h}}_{v}^{\ell-1}\right)^{\top}\left(\boldsymbol{W}_{\ell}^{K} \overline{\boldsymbol{h}}_{u}^{\ell-1}\right) / \sqrt{d_{\mathrm{TF}}} \tag{1} $$ ...

March 30, 2022 · 1 min · JhuoW

NeurIPS2021 《Not All Low-Pass Filters are Robust in Graph Convolutional Networks》 Reading Notes

paper Introduction 很多GNN易受图结构攻击的影响,本文首先证明了symmetric normalized laplacian的低频分量作为GCN的filter,在某个特征值区间内对于结构扰动更加robust。基于该理论提出GCN-LFR,通过一个辅助神经网络迁移低频分量的robustness。 Q: 对抗扰动边是否会对graph spectrum产生同等的影响?过去的研究显示来自结构攻击的扰动在图谱上表达了一种隐含的趋势。如下图所示,结构扰动后,小的特征值(低频)变化较小, 高频变化较大,即高频对扰动更加敏感。 本文证明了当normalized symmetric laplacian的特征值落于某个特定区间时,低频分量会更加robust。 Methodology Poisoning Attack是指训练前扰动: Problem 1 (Poisoning Attack): 给定一个扰动图 $\mathcal{G}^\prime$, 要对目标集合$\mathcal{T}$做对抗防御的目的是设计一个更加鲁棒的模型,使得模型在扰动图上训练后对$\mathcal{T}$中节点的预测结果和在原图上训练得到的预测结果相似: $$ \min_{\boldsymbol{\Theta}^{r *}} \sum_{u \in \mathcal{T}}\left|\left|\mathcal{M}_{u}^{r}\left(\boldsymbol{A}^{\prime}, \boldsymbol{X} ; \boldsymbol{\Theta}^{r *}\right)-\mathcal{M}_{u}\left(\boldsymbol{A}, \boldsymbol{X} ; \boldsymbol{\Theta}^{*}\right)\right|\right| $$ 即模型可以尽可能避免扰动对预测带来的影响。 其中$\mathcal{M}_{u}^{r}\left(\boldsymbol{A}^{\prime}, \boldsymbol{X} ; \boldsymbol{\Theta}^{r *}\right)=\hat{\boldsymbol{y}}_{u}^{r}$模型在扰动图上训练过后对节点$u$的预测,$\mathcal{M}_{u}^{r}$是在扰动图上训练过后的模型,最佳参数为$\boldsymbol{\Theta}^{r *}$, $\mathcal{M}$是在原图上训练过后的模型。 观察$\hat{\boldsymbol{A}}$的特征值, 因为$\hat{\boldsymbol{A}} = I - L$, 所以$\hat{\boldsymbol{A}}$的大特征值对应于低频分量,小特征值对应于高频分量。 从图1可以看出$\hat{\boldsymbol{A}}$的大特征值对于结构扰动更加鲁棒, 因为扰动之后大特征值的变化较小。所以GCN-SVD只是用最低频的分量来做defense. 接下来本文证明了只有一条边被扰动的情况下,一定存在低频$\lambda_b$,比所有高频都robust, 鲁棒区间为: $$ \max \left(0, \frac{d_{b}-d_{a}+c_{a} \lambda_{a}}{c_{b}}\right)<\lambda_{b}<\min \left(\frac{d_{b}+d_{a}-c_{a} \lambda_{a}}{c_{b}}, 1\right) $$ 即,当特征值落于这个区间中时,它一定比高频更加鲁棒。 对于Non-targeted Perturbation, 图中有多条边被扰动,那么特征值的鲁棒区间为: $$ \max_{v \in \mathcal{P}_{u}, u \in \mathcal{T}}\left(0, \frac{d_{b u v}-d_{a u v}+c_{a u v} \lambda_{a}}{c_{b u v}}\right)<\lambda_{b}<\min_{v \in \mathcal{P}_{u}, u \in \mathcal{T}}\left(\frac{d_{b u v}+d_{a u v}-c_{a u v} \lambda_{a}}{c_{b u v}}, 1\right) $$ 在得到不同扰动情况下的鲁棒区间后,我们可以基于鲁棒区间来增强GCN的鲁棒性。 ...

March 29, 2022 · 1 min · JhuoW

Blog, Tools and Survey

这篇笔记用于收藏别人的博客 Tech Blog Blog Author https://michael-bronstein.medium.com/ Michael Bronstein https://geometricdeeplearning.com/ Michael Bronstein https://www.notion.so/Paper-Notes-by-Vitaly-Kurin-97827e14e5cd4183815cfe3a5ecf2f4c Vitaly Kurin (Many Paper Notes) https://uvadlc-notebooks.readthedocs.io/en/latest/tutorial_notebooks/tutorial1/Lisa_Cluster.html UvA DL Notebooks https://graph-neural-networks.github.io/index.html GNN Books http://prob140.org/sp17/textbook/ Probability for Data Science class at UC Berkeley https://graphreason.github.io/schedule.html Learning and Reasoning with Graph-Structured Representations ICML 2019 Workshop https://chuxuzhang.github.io/KDD21_Tutorial.html KDD2021 Tutorial: Data Efficient Learning on Graphs http://songcy.net/posts/ Changyue Song (Kernel) https://www.cs.mcgill.ca/~wlh/grl_book/ William L. Hamilton https://kexue.fm/ BoJone https://danielegrattarola.github.io/blog/ Daniele Grattarola (EPFL) https://ai.googleblog.com/2022/03/robust-graph-neural-networks.html Google AI Blog https://zhiyuchen.com/blogs/ Zhiyu Chen https://andreasloukas.blog/ Andreas Loukas (EPFL) https://irhum.pubpub.org/pub/gnn/release/4 Understanding Graph Neural Networks https://lilianweng.github.io/ Lilian Weng https://www.zhihu.com/column/marlin 深度学习与图网络 https://github.com/roboticcam/machine-learning-notes Yida Xu https://www.dgl.ai/pages/index.html DGL https://www.kexinhuang.com/tech-blog Kexin Huang https://rish16.notion.site/a8c6fcb0c29c4764afa1926ad33084f8?v=bb27bb0444574fbd85f0c9d7e43b9da8 Rishabh Anand https://saashanair.com/blog Saasha Nair http://www.huaxiaozhuan.com/ 华校专 https://github.com/dglai/WWW20-Hands-on-Tutorial DGL https://blog.csdn.net/CSDNTianJi/article/details/104195306 Meng Liu https://www.chaitjo.com/post/ Chaitanya K. Joshi https://scottfreitas.medium.com/ Scott Freitas https://fabianfuchsml.github.io/ Fabian Fuchs https://medium.com/@pantelis.elinas Pantelis Elinas https://github.com/tianyicui/pack 背包9講 https://www.fenghz.xyz/ https://sakigami-yang.me/2017/08/13/about-kernel-01/ kernel https://davidham3.github.io/blog https://fenghz.github.io/index.html https://archwalker.github.io/ Awesome-Awesomes Repo Name https://github.com/naganandy/graph-based-deep-learning-literature links to conference publications in graph-based deep learning (Very, Very, Very Important) https://github.com/SherylHYX/pytorch_geometric_signed_directed PyTorch Geometric Signed Directed is a signed/directed graph neural network extension library for PyTorch Geometric. https://github.com/EdisonLeeeee/Awesome-Fair-Graph-Learning Paper Lists for Fair Graph Learning https://github.com/thunlp/PromptPapers Must-read papers on prompt-based tuning for pre-trained language models. https://github.com/zhao-tong/graph-data-augmentation-papers A curated list of graph data augmentation papers. https://github.com/Thinklab-SJTU/ThinkMatch Code & pretrained models of novel deep graph matching methods. https://github.com/FLHonker/Awesome-Knowledge-Distillation Awesome Knowledge-Distillation. 分类整理的知识蒸馏paper(2014-2021)。 https://github.com/zlpure/awesome-graph-representation-learning A curated list for awesome graph representation learning resources. https://github.com/basiralab/GNNs-in-Network-Neuroscience A review of papers proposing novel GNN methods with application to brain connectivity published in 2017-2020. https://github.com/flyingdoog/awesome-graph-explainability-papers Papers about explainability of GNNs https://github.com/yuanqidu/awesome-graph-generation A curated list of graph generation papers and resources. https://github.com/benedekrozemberczki/awesome-decision-tree-papers A collection of research papers on decision, classification and regression trees with implementations. https://github.com/AstraZeneca/awesome-explainable-graph-reasoning A collection of research papers and software related to explainability in graph machine learning. https://github.com/LirongWu/awesome-graph-self-supervised-learning Awesome Graph Self-Supervised Learning https://github.com/Chen-Cai-OSU/awesome-equivariant-network Paper list for equivariant neural network https://github.com/mengliu1998/DL4DisassortativeGraphs Papers about developing DL methods on disassortative graphs https://github.com/SunQingYun1996/Graph-Reinforcement-Learning-Papers A curated list of graph reinforcement learning papers. https://github.com/ChandlerBang/awesome-self-supervised-gnn Papers about pretraining and self-supervised learning on Graph Neural Networks (GNN). https://github.com/GRAND-Lab/Awesome-Graph-Neural-Networks Paper Lists for Graph Neural Networks https://github.com/jwzhanggy/IFMLab_GNN Graph Neural Network Models from IFM Lab https://github.com/ChandlerBang/awesome-graph-attack-papers Adversarial attacks and defenses on Graph Neural Networks. https://github.com/safe-graph/graph-adversarial-learning-literature A curated list of adversarial attacks and defenses papers on graph-structured data. https://github.com/benedekrozemberczki/awesome-graph-classification A collection of important graph embedding, classification and representation learning papers with implementations. https://github.com/benedekrozemberczki/awesome-gradient-boosting-papers A curated list of gradient boosting research papers with implementations. https://github.com/benedekrozemberczki/awesome-community-detection A curated list of community detection research papers with implementations. https://github.com/giannifranchi/awesome-uncertainty-deeplearning This repository contains a collection of surveys, datasets, papers, and codes, for predictive uncertainty estimation in deep learning models. https://sites.google.com/site/graphmatchingmethods/ Efficient Methods for Graph Matching and MAP Inference https://github.com/yueliu1999/Awesome-Deep-Graph-Clustering Awesome Deep Graph Clustering is a collection of SOTA, novel deep graph clustering methods (papers, codes, and datasets). https://github.com/jwwthu/GNN4Traffic This is the repository for the collection of Graph Neural Network for Traffic Forecasting. https://github.com/zwt233/Awesome-Auto-GNNs A paper collection about automated graph learning https://github.com/chaitjo/awesome-efficient-gnn Efficient Graph Neural Networks - a curated list of papers and projects https://github.com/Radical3-HeZhang/Awesome-Trustworthy-GNNs Awesome Resources on Trustworthy Graph Neural Networks https://github.com/EdisonLeeeee/Awesome-Masked-Autoencoders A collection of literature after or concurrent with Masked Autoencoder (MAE) (Kaiming He el al.). https://github.com/XiaoxiaoMa-MQ/Awesome-Deep-Graph-Anomaly-Detection Awesome graph anomaly detection techniques built based on deep learning frameworks. https://github.com/mengliu1998/awesome-expressive-gnn A collection of papers studying/improving the expressiveness of graph neural networks (GNNs) Useful Repo/Tools Name Info http://acronymify.com/ Model Name https://csacademy.com/app/graph_editor/ Graph Editor https://github.com/guanyingc/python_plot_utils A simple code for plotting figure, colorbar, and cropping with python https://github.com/guanyingc/latex_paper_writing_tips Tips for Writing a Research Paper using LaTeX https://github.com/JhuoW/Pytorch_Program_Templete Pytorch Program Templete GNN https://github.com/graph4ai/graph4nlp Graph4nlp is the library for the easy use of Graph Neural Networks for NLP. Welcome to visit our DLG4NLP website (https://dlg4nlp.github.io/index.html) for various learning resources! https://github.com/benedekrozemberczki/pytorch_geometric_temporal PyTorch Geometric Temporal: Spatiotemporal Signal Processing with Neural Machine Learning Models (CIKM 2021) https://github.com/ysig/GraKeL A scikit-learn compatible library for graph kernels https://github.com/jajupmochi/graphkit-learn A python package for graph kernels, graph edit distances, and graph pre-image problem. https://github.com/pliang279/awesome-phd-advice Collection of advice for prospective and current PhD students https://github.com/MLEveryday/100-Days-Of-ML-Code 100-Days-Of-ML-Code中文版 https://github.com/d2l-ai/d2l-zh 《动手学深度学习》 https://github.com/lukas-blecher/LaTeX-OCR pix2tex: Using a ViT to convert images of equations into LaTeX code. https://github.com/thunlp/OpenPrompt An Open-Source Framework for Prompt-Learning. https://github.com/snap-stanford/GraphGym Platform for designing and evaluating Graph Neural Networks (GNN) https://github.com/pygod-team/pygod A Python Library for Graph Outlier Detection (Anomaly Detection) https://github.com/MLNLP-World/Paper_Writing_Tips latex写作建议 https://github.com/dair-ai/ML-YouTube-Courses A place to discover the latest machine learning courses on YouTube. Miscellaneous Name Desc https://github.com/The-Run-Philosophy-Organization/run run学指南 https://10beasts.net/ 测评

March 29, 2022 · 4 min · JhuoW