[init] re-initialize the project by update the structure

Omniscient-KG · Apr 18, 2023 · 5404834 · 5404834
1 parent b518c01
commit 5404834
Show file tree

Hide file tree

Showing 70 changed files with 166 additions and 170 deletions.
diff --git a/Applications/Ads Recommendation/Click-Through Rate/DLReview.md b/Applications/Ads Recommendation/Click-Through Rate/DLReview.md
@@ -0,0 +1,18 @@
+## Deep Learning for Click-Through Rate Estimation
+
+### 0. 概括
+
+本文是一篇关于深度学习在CTR预估中的综述，文章介绍了CTR预估的重要性以及深度学习模型如何提高CTR预估的性能。
+
+
+### 1. CTR背景
+
+CTR预估是指通过分析用户的历史行为数据，预测用户是否会点击某个广告或者某个推荐内容。CTR预估任务的数据案例如下图所示：
+
+<div align="center">
+<img src=Figure/ClickInstance.png width=75% />
+</div>
+
+CTR预估的模型训练任务可构建为二分类问题，且损失函数如下：
+$$ \mathcal{L}(x, y, \theta) = -y\cdot \log \sigma(f_{\theta}(x)) - (1-y)\cdot \log (1-\sigma(f_{\theta} (x))) $$
+
diff --git a/Applications/Ads Recommendation/Click-Through Rate/Figure/ClickInstance.png b/Applications/Ads Recommendation/Click-Through Rate/Figure/ClickInstance.png
diff --git a/Applications/Computer Vision/README.md → Applications/Ads Recommendation/basicNote.md b/Applications/Computer Vision/README.md → Applications/Ads Recommendation/basicNote.md
diff --git a/...ons/Natural Language Processing/README.md → Applications/Computer Vision/basicNote.md b/...ons/Natural Language Processing/README.md → Applications/Computer Vision/basicNote.md
diff --git a/Applications/Programming Language/README.md → .../Natural Language Processing/basicNote.md b/Applications/Programming Language/README.md → .../Natural Language Processing/basicNote.md
diff --git a/Applications/Robotics/README.md → ...cations/Programming Language/basicNote.md b/Applications/Robotics/README.md → ...cations/Programming Language/basicNote.md
diff --git a/...ine Learning/Generative Model/Beta-VAE.md → Applications/Robotics/basicNote.md b/...ine Learning/Generative Model/Beta-VAE.md → Applications/Robotics/basicNote.md
diff --git a/...ine Learning/Bayesian Learning/Dropout.md → ...ne Learning/Bayesian Inference/Dropout.md b/...ine Learning/Bayesian Learning/Dropout.md → ...ne Learning/Bayesian Inference/Dropout.md
diff --git a/Machine Learning/Generative Model/CVAE.md → ...enerative AI/Generative Model/Beta-VAE.md b/Machine Learning/Generative Model/CVAE.md → ...enerative AI/Generative Model/Beta-VAE.md
diff --git a/Machine Learning/Generative Model/DDPM.md → ...ng/Generative AI/Generative Model/CVAE.md b/Machine Learning/Generative Model/DDPM.md → ...ng/Generative AI/Generative Model/CVAE.md
diff --git a/...e Learning/Generative Model/ControlVAE.md → ...erative AI/Generative Model/ControlVAE.md b/...e Learning/Generative Model/ControlVAE.md → ...erative AI/Generative Model/ControlVAE.md
diff --git a/Machine Learning/Generative Model/GAN.md → ...ng/Generative AI/Generative Model/DDPM.md b/Machine Learning/Generative Model/GAN.md → ...ng/Generative AI/Generative Model/DDPM.md
diff --git a/Machine Learning/Generative Model/LDM.md → ...ing/Generative AI/Generative Model/GAN.md b/Machine Learning/Generative Model/LDM.md → ...ing/Generative AI/Generative Model/GAN.md
diff --git a/Machine Learning/Generative Model/VAE.md → ...ing/Generative AI/Generative Model/LDM.md b/Machine Learning/Generative Model/VAE.md → ...ing/Generative AI/Generative Model/LDM.md
diff --git a/Machine Learning/Generative Model/VQ-GAN.md → ...ing/Generative AI/Generative Model/VAE.md b/Machine Learning/Generative Model/VQ-GAN.md → ...ing/Generative AI/Generative Model/VAE.md
diff --git a/Machine Learning/Generative Model/VQ-VAE.md → .../Generative AI/Generative Model/VQ-GAN.md b/Machine Learning/Generative Model/VQ-VAE.md → .../Generative AI/Generative Model/VQ-GAN.md
diff --git a/Machine Learning/Graph Learning/GAT.md → .../Generative AI/Generative Model/VQ-VAE.md b/Machine Learning/Graph Learning/GAT.md → .../Generative AI/Generative Model/VQ-VAE.md
diff --git a/...earning as Implicit Bayesian Inference.md → ...earning as Implicit Bayesian Inference.md b/...earning as Implicit Bayesian Inference.md → ...earning as Implicit Bayesian Inference.md
diff --git a/Machine Learning/Pre-training Model/BERT.md → .../Generative AI/Pre-training Model/BERT.md b/Machine Learning/Pre-training Model/BERT.md → .../Generative AI/Pre-training Model/BERT.md
@@ -16,7 +16,7 @@ BERT是一种自监督的语言模型，通过利用大量未标注的语料库
 > #### *模型架构*
 > Bert模型架构是多层双向Transformer Encoder。
 
-![Model Structure Comparison](./Figure/Bert%20Structure.png)
+![Model Structure Comparison](Figure/Bert%20Structure.png)
 
 
 
@@ -43,7 +43,7 @@ BERT是一种自监督的语言模型，通过利用大量未标注的语料库
 > 下图展示了Bert的四种微调任务，（a）和（b）是sentence-level的任务，（c）和（d）token-level的任务。
 > 
 
-![Bert Fine-tuning](./Figure/Bert%20fine-tuning.png)
+![Bert Fine-tuning](Figure/Bert%20fine-tuning.png)
 
 
 ### 2. 评估任务

diff --git a/...-training Model/Figure/Bert Structure.png → ...-training Model/Figure/Bert Structure.png b/...-training Model/Figure/Bert Structure.png → ...-training Model/Figure/Bert Structure.png
diff --git a/...raining Model/Figure/Bert fine-tuning.png → ...raining Model/Figure/Bert fine-tuning.png b/...raining Model/Figure/Bert fine-tuning.png → ...raining Model/Figure/Bert fine-tuning.png
diff --git a/Machine Learning/Pre-training Model/GPT.md → ...g/Generative AI/Pre-training Model/GPT.md b/Machine Learning/Pre-training Model/GPT.md → ...g/Generative AI/Pre-training Model/GPT.md
diff --git a/Machine Learning/Pre-training Model/GPT2.md → .../Generative AI/Pre-training Model/GPT2.md b/Machine Learning/Pre-training Model/GPT2.md → .../Generative AI/Pre-training Model/GPT2.md
diff --git a/Machine Learning/Pre-training Model/GPT3.md → .../Generative AI/Pre-training Model/GPT3.md b/Machine Learning/Pre-training Model/GPT3.md → .../Generative AI/Pre-training Model/GPT3.md
diff --git a/...earning/Pre-training Model/InstructGPT.md → ...tive AI/Pre-training Model/InstructGPT.md b/...earning/Pre-training Model/InstructGPT.md → ...tive AI/Pre-training Model/InstructGPT.md
diff --git a/Machine Learning/Pre-training Model/LaMDA.md → ...Generative AI/Pre-training Model/LaMDA.md b/Machine Learning/Pre-training Model/LaMDA.md → ...Generative AI/Pre-training Model/LaMDA.md
diff --git a/...ine Learning/Pre-training Model/WebGPT.md → ...enerative AI/Pre-training Model/WebGPT.md b/...ine Learning/Pre-training Model/WebGPT.md → ...enerative AI/Pre-training Model/WebGPT.md
diff --git a/...ne Learning/Reinforcement Learning/DQN.md → ...nerative AI/Reinforcement Learning/DQN.md b/...ne Learning/Reinforcement Learning/DQN.md → ...nerative AI/Reinforcement Learning/DQN.md
diff --git a/...ne Learning/Reinforcement Learning/PPO.md → ...nerative AI/Reinforcement Learning/PPO.md b/...ne Learning/Reinforcement Learning/PPO.md → ...nerative AI/Reinforcement Learning/PPO.md
diff --git a/...rning/Reinforcement Learning/basicNote.md → ...ve AI/Reinforcement Learning/basicNote.md b/...rning/Reinforcement Learning/basicNote.md → ...ve AI/Reinforcement Learning/basicNote.md
diff --git a/Machine Learning/Graph Learning/GCN.md → ...raph-Network/Graph Neural Networks/GAT.md b/Machine Learning/Graph Learning/GCN.md → ...raph-Network/Graph Neural Networks/GAT.md
diff --git a/Machine Learning/Graph Learning/GTN.md → ...raph-Network/Graph Neural Networks/GCN.md b/Machine Learning/Graph Learning/GTN.md → ...raph-Network/Graph Neural Networks/GCN.md
diff --git a/Machine Learning/Graph Learning/HAN.md → ...raph-Network/Graph Neural Networks/GTN.md b/Machine Learning/Graph Learning/HAN.md → ...raph-Network/Graph Neural Networks/GTN.md
diff --git a/Machine Learning/Graph Learning/GraphSage.md → ...etwork/Graph Neural Networks/GraphSage.md b/Machine Learning/Graph Learning/GraphSage.md → ...etwork/Graph Neural Networks/GraphSage.md
diff --git a/Machine Learning/Graph Learning/R-GCNs.md → ...raph-Network/Graph Neural Networks/HAN.md b/Machine Learning/Graph Learning/R-GCNs.md → ...raph-Network/Graph Neural Networks/HAN.md
diff --git a/Machine Learning/Graph Learning/TransD.md → ...h-Network/Graph Neural Networks/R-GCNs.md b/Machine Learning/Graph Learning/TransD.md → ...h-Network/Graph Neural Networks/R-GCNs.md
diff --git a/... Learning/Graph Learning/AnonymousWalk.md → ... Representation Learning/AnonymousWalk.md b/... Learning/Graph Learning/AnonymousWalk.md → ... Representation Learning/AnonymousWalk.md
@@ -55,5 +55,5 @@ $$ y(w_t) = b + Uh(w_{t-\Delta}, \cdots, w_{t+\Delta}, d) $$
 其中$b\in R$和$U \in R^{d_a + d_g}$是Softmax参数，$h$计算可参见下图，先对匿名游走结果的嵌入向量$(w_{t-\Delta}, \cdots, w_{t+\Delta})$进行平均，再和图嵌入$d$进行联结。
 
 <div align="center">
-<img src=./Figure/AWE.png width=40% />
+<img src=Figure/AWE.png width=40% />
 </div>
diff --git a/Machine Learning/Graph Learning/DeepWalk.md → ...Graph Representation Learning/DeepWalk.md b/Machine Learning/Graph Learning/DeepWalk.md → ...Graph Representation Learning/DeepWalk.md
@@ -38,7 +38,7 @@ $$\mathrm{minimize} - \log Pr(\begin{Bmatrix} v_{i-w}, \cdots, v_{i-1}, v_{i+1},
 ### 4. DeepWalk方法详情
 
 <div align="center">
-<img src=./Figure/OverviewDeepWalk.png width=60% />
+<img src=Figure/OverviewDeepWalk.png width=60% />
 </div>
 
  DeepWalk的整体流程如上图所示，其伪代码如下图所示。
@@ -50,11 +50,11 @@ $$\mathrm{minimize} - \log Pr(\begin{Bmatrix} v_{i-w}, \cdots, v_{i-1}, v_{i+1},
  其中$\Psi(b_l)\in \mathbb{R}^{d}$是节点$b_l$的父节点表征，$\Phi$和$\Psi$为待优化的模型参数。
 
 <div align="center">
-<img src=./Figure/DeepWalk.png width=40% />
+<img src=Figure/DeepWalk.png width=40% />
 </div>
 
 <div align="center">
-<img src=./Figure/SkipGram.png width=40% />
+<img src=Figure/SkipGram.png width=40% />
 </div>
 
 
@@ -76,5 +76,5 @@ $$\mathrm{minimize} - \log Pr(\begin{Bmatrix} v_{i-w}, \cdots, v_{i-1}, v_{i+1},
  评估结果如下图所示
 
 <div align="center">
-<img src=./Figure/DeepWalkEval.png width=40% />
+<img src=Figure/DeepWalkEval.png width=40% />
 </div>
diff --git a/...ne Learning/Graph Learning/Figure/AWE.png → ...ph Representation Learning/Figure/AWE.png b/...ne Learning/Graph Learning/Figure/AWE.png → ...ph Representation Learning/Figure/AWE.png
diff --git a/...arning/Graph Learning/Figure/DeepWalk.png → ...presentation Learning/Figure/DeepWalk.png b/...arning/Graph Learning/Figure/DeepWalk.png → ...presentation Learning/Figure/DeepWalk.png
diff --git a/...ng/Graph Learning/Figure/DeepWalkEval.png → ...entation Learning/Figure/DeepWalkEval.png b/...ng/Graph Learning/Figure/DeepWalkEval.png → ...entation Learning/Figure/DeepWalkEval.png
diff --git a/...arning/Graph Learning/Figure/Graphlet.png → ...presentation Learning/Figure/Graphlet.png b/...arning/Graph Learning/Figure/Graphlet.png → ...presentation Learning/Figure/Graphlet.png
diff --git a/...earning/Graph Learning/Figure/LINEval.png → ...epresentation Learning/Figure/LINEval.png b/...earning/Graph Learning/Figure/LINEval.png → ...epresentation Learning/Figure/LINEval.png
diff --git a/...raph Learning/Figure/OverviewDeepWalk.png → ...tion Learning/Figure/OverviewDeepWalk.png b/...raph Learning/Figure/OverviewDeepWalk.png → ...tion Learning/Figure/OverviewDeepWalk.png
diff --git a/...arning/Graph Learning/Figure/SkipGram.png → ...presentation Learning/Figure/SkipGram.png b/...arning/Graph Learning/Figure/SkipGram.png → ...presentation Learning/Figure/SkipGram.png
diff --git a/...h Learning/Figure/SpectralClusteringN.png → ...n Learning/Figure/SpectralClusteringN.png b/...h Learning/Figure/SpectralClusteringN.png → ...n Learning/Figure/SpectralClusteringN.png
diff --git a/... Learning/Figure/SpectralClusteringN2.png → ... Learning/Figure/SpectralClusteringN2.png b/... Learning/Figure/SpectralClusteringN2.png → ... Learning/Figure/SpectralClusteringN2.png
diff --git a/...h Learning/Figure/SpectralClusteringU.png → ...n Learning/Figure/SpectralClusteringU.png b/...h Learning/Figure/SpectralClusteringU.png → ...n Learning/Figure/SpectralClusteringU.png
diff --git a/...Learning/Graph Learning/Figure/TransE.png → ...Representation Learning/Figure/TransE.png b/...Learning/Graph Learning/Figure/TransE.png → ...Representation Learning/Figure/TransE.png
diff --git a/...g/Graph Learning/Figure/TransH-TransE.png → ...ntation Learning/Figure/TransH-TransE.png b/...g/Graph Learning/Figure/TransH-TransE.png → ...ntation Learning/Figure/TransH-TransE.png
diff --git a/...ng/Graph Learning/Figure/TransHmodels.png → ...entation Learning/Figure/TransHmodels.png b/...ng/Graph Learning/Figure/TransHmodels.png → ...entation Learning/Figure/TransHmodels.png
diff --git a/Machine Learning/Graph Learning/LINE.md → ...ork/Graph Representation Learning/LINE.md b/Machine Learning/Graph Learning/LINE.md → ...ork/Graph Representation Learning/LINE.md
@@ -44,5 +44,5 @@
 ### 3. 评估
 
 <div align="center">
-<img src=./Figure/LINEval.png width=40% />
+<img src=Graph width=40% />
 </div>
diff --git a/...ning/Graph Learning/SpectralClustering.md → ...esentation Learning/SpectralClustering.md b/...ning/Graph Learning/SpectralClustering.md → ...esentation Learning/SpectralClustering.md
@@ -42,17 +42,17 @@ $$ f^{\prime} L_{sym} f = \frac{1}{2} \sum_{i, j=1}^{n} w_{i, j} \left( \frac{f_
  $$ s(x_i, x_j) = e^{-\frac{|x_i - x_j|^2}{2 \sigma^2}} $$
 
 <div align="center">
-<img src=./Figure/SpectralClusteringU.png width=40% />
+<img src=Graph width=40% />
 </div>
 
  归一化的谱聚类方法如下：
 
 <div align="center">
-<img src=./Figure/SpectralClusteringN.png width=40% />
+<img src=Graph width=40% />
 </div>
 
 <div align="center">
-<img src=./Figure/SpectralClusteringN2.png width=40% />
+<img src=Graph width=40% />
 </div>
 
 

diff --git a/Machine Learning/Graph Learning/TransR.md → ...k/Graph Representation Learning/TransD.md b/Machine Learning/Graph Learning/TransR.md → ...k/Graph Representation Learning/TransD.md
diff --git a/Machine Learning/Graph Learning/TransE.md → ...k/Graph Representation Learning/TransE.md b/Machine Learning/Graph Learning/TransE.md → ...k/Graph Representation Learning/TransE.md
@@ -25,5 +25,5 @@ $$  S^{\prime}_{(h, l, t)} = \left\lbrace (h^{\prime}, l, t) | h^{\prime} \in E
 在优化过程中，实体嵌入向量的L2正则约束为1，关系嵌入向量没有约束。详细地优化过程如下图所示。
 
 <div align="center">
-<img src=./Figure/TransE.png width=40% />
+<img src=Graph width=40% />
 </div>
diff --git a/Machine Learning/Graph Learning/TransH.md → ...k/Graph Representation Learning/TransH.md b/Machine Learning/Graph Learning/TransH.md → ...k/Graph Representation Learning/TransH.md
@@ -10,7 +10,7 @@ TransE难以处理自映射、一对多、多对多等不同属性的关系，
 前人方法和TransH的Score函数和参数量级对比结果如下图所示。
 
 <div align="center">
-<img src=./Figure/TransHmodels.png width=40% />
+<img src=Graph width=40% />
 </div>
 
 
@@ -21,7 +21,7 @@ TransE难以处理自映射、一对多、多对多等不同属性的关系，
 记$\Delta$是存在关系的三元组集合，$(h, r, t)\in \Delta$表示三元组关系是存在的。
 
 <div align="center">
-<img src=./Figure/TransH-TransE.png width=40% />
+<img src=Graph width=40% />
 </div>
 
 

diff --git a/...earning/Optimization Algorithm/RMSProp.md → ...k/Graph Representation Learning/TransR.md b/...earning/Optimization Algorithm/RMSProp.md → ...k/Graph Representation Learning/TransR.md
diff --git a/Machine Learning/Graph Learning/graph2vec.md → ...raph Representation Learning/graph2vec.md b/Machine Learning/Graph Learning/graph2vec.md → ...raph Representation Learning/graph2vec.md
diff --git a/Machine Learning/Graph Learning/graphlet.md → ...Graph Representation Learning/graphlet.md b/Machine Learning/Graph Learning/graphlet.md → ...Graph Representation Learning/graphlet.md
@@ -10,7 +10,7 @@
 大小为$k$的graphlet可表示为$\mathcal{G} = \begin{Bmatrix} graphlet(1), \cdots, graphlet(N_k) \end{Bmatrix}$，大小为4的所有Graphlet如下图所示。
 
 <div align="center">
-<img src=./Figure/Graphlet.png width=40% />
+<img src=Figure/Graphlet.png width=40% />
 </div>
 
 图G的k谱$f_G$是图中大小为$k$的graphlet的子图数量：$f_G (i) = \mathrm{number}(graphlet(i) \subseteq G)$，归一化后的向量为：

diff --git a/Machine Learning/Graph Learning/basicNote.md → Machine Learning/Graph-Network/basicNote.md b/Machine Learning/Graph Learning/basicNote.md → Machine Learning/Graph-Network/basicNote.md
diff --git a/...e Learning/Active Learning/NbrGraphSGD.md → ...g Paradigm/Active Learning/NbrGraphSGD.md b/...e Learning/Active Learning/NbrGraphSGD.md → ...g Paradigm/Active Learning/NbrGraphSGD.md
diff --git a/...ning/Semi-Supervised Learning/AdaMatch.md → ...digm/Semi-Supervised Learning/AdaMatch.md b/...ning/Semi-Supervised Learning/AdaMatch.md → ...digm/Semi-Supervised Learning/AdaMatch.md
diff --git a/...timization Algorithm/Figure/optimizer.gif → ...mization/GradientOpt/Figure/optimizer.gif b/...timization Algorithm/Figure/optimizer.gif → ...mization/GradientOpt/Figure/optimizer.gif
diff --git a/...arning/Optimization Algorithm/momentum.md → ...rning/Optimization/GradientOpt/RMSProp.md b/...arning/Optimization Algorithm/momentum.md → ...rning/Optimization/GradientOpt/RMSProp.md
diff --git a/...ne Learning/Optimization Algorithm/SGD.md → ... Learning/Optimization/GradientOpt/SGD.md b/...ne Learning/Optimization Algorithm/SGD.md → ... Learning/Optimization/GradientOpt/SGD.md
diff --git a/...earning/Optimization Algorithm/adaGrad.md → ...rning/Optimization/GradientOpt/adaGrad.md b/...earning/Optimization Algorithm/adaGrad.md → ...rning/Optimization/GradientOpt/adaGrad.md
diff --git a/...e Learning/Optimization Algorithm/adam.md → ...Learning/Optimization/GradientOpt/adam.md b/...e Learning/Optimization Algorithm/adam.md → ...Learning/Optimization/GradientOpt/adam.md
diff --git a/...rning/Optimization Algorithm/basisNote.md → ...ing/Optimization/GradientOpt/basisNote.md b/...rning/Optimization Algorithm/basisNote.md → ...ing/Optimization/GradientOpt/basisNote.md
@@ -1,5 +1,5 @@
 <div align="center">
-<img src=./Figure/optimizer.gif width=60% />
+<img src=Figure/optimizer.gif width=60% />
 </div>
 
 

diff --git a/Machine Learning/README.md → ...ning/Optimization/GradientOpt/momentum.md b/Machine Learning/README.md → ...ning/Optimization/GradientOpt/momentum.md