Concept List¶
Total concepts: 300
This list feeds the learning graph dependency DAG. Review before proceeding.
Prerequisites¶
- Matrix Multiplication
- Matrix Transpose
- Eigenvalue Decomposition
- Eigenvector
- Symmetric Matrix
- Positive Semi-Definite Matrix
- Singular Value Decomp (SVD)
- Matrix Rank
- Dot Product
- Cosine Similarity
- Gradient Descent
- Backpropagation
- Chain Rule (Calculus)
- Automatic Differentiation
- PyTorch Tensor
- PyTorch Autograd
- Neural Network Layer
- Softmax Function
- Cross-Entropy Loss
- Adam Optimizer
Graph Fundamentals¶
- Graph (Undirected)
- Graph (Directed)
- Node (Vertex)
- Edge (Link)
- Adjacency Matrix
- Node Degree
- In-Degree
- Out-Degree
- Degree Distribution
- Power-Law Degree Distribution
- Graph Path
- Shortest Path
- Graph Diameter
- Connected Component
- Strongly Connected Component
- Weakly Connected Component
- Bipartite Graph
- Heterogeneous Graph
- Multigraph
- Weighted Graph
- Attribute Graph
- Subgraph
- Ego Network
- Clique
- Cycle
- Tree (Graph Theory)
- Spanning Tree
- Planar Graph
- Graph Isomorphism
- Graph Homomorphism
- Small-World Network
- Scale-Free Network
- Erdős–Rényi Random Graph
- Barabási–Albert Model
- Preferential Attachment
- Giant Component
- Transitivity
Classical Graph Algorithms¶
- Local Clustering Coefficient
- Betweenness Centrality
- Closeness Centrality
- Eigenvector Centrality
- PageRank
- Personalized PageRank
- HITS Algorithm
- Random Walk
- Stationary Distribution
- Power Iteration
- Teleportation (PageRank)
- Community Detection
- Modularity (Network)
- Louvain Algorithm
- Girvan-Newman Algorithm
- Spectral Clustering
- Normalized Cut
- Overlapping Community
- BigCLAM Model
- Network Motif
- Graphlet
- Graphlet Degree Vector
- Graph Kernel
- Weisfeiler-Lehman Kernel
- Label Propagation
- Belief Propagation
- Influence Maximization
- Linear Threshold Model
- Independent Cascade Model
- SIR Epidemic Model
- Breadth-First Search
- Depth-First Search
- Katz Similarity
Node Embeddings¶
- Node Embedding
- Embedding Space
- Encoder-Decoder Framework
- Shallow Embedding
- Matrix Factorization (Graph)
- DeepWalk
- node2vec
- Biased Random Walk
- BFS Strategy (node2vec)
- DFS Strategy (node2vec)
- Skip-Gram Model
- Negative Sampling
- LINE Embedding
- Transductive Learning
- Inductive Learning
- Structural Equivalence
- Homophily
GNN Architecture¶
- Graph Neural Network (GNN)
- Message Passing Neural Net
- Message Function
- Aggregation Function
- Update Function
- Graph Convolutional Network
- GraphSAGE
- Graph Attention Network (GAT)
- Attention Mechanism (Graph)
- Multi-Head Attention (Graph)
- Graph Isomorphism Network
- Sum Aggregation
- Mean Aggregation
- Max Aggregation
- Neighborhood Aggregation
- K-Hop Neighborhood
- Receptive Field (GNN)
- Layer Depth (GNN)
- Skip Connection (GNN)
- Residual Connection (GNN)
- Jumping Knowledge Network
- Graph-Level Readout
- Global Mean Pooling
- Global Sum Pooling
- DiffPool
- MinCutPool
- Node-Level Task
- Edge-Level Task
- Graph-Level Task
- Link Prediction
- Node Classification
- Graph Classification
- Graph Regression
- Spectral Graph Convolution
- Chebyshev Polynomial Conv
- Graph Laplacian
- Normalized Graph Laplacian
- Spectral Domain (Graph)
- Spatial Domain (Graph)
- Virtual Node Augmentation
- Virtual Edge Augmentation
GNN Theory¶
- Weisfeiler-Lehman (WL) Test
- 1-WL Test
- k-WL Test
- GNN Expressiveness
- Graph Isomorphism Problem
- GNN Distinguishing Power
- Over-Smoothing
- Over-Squashing
- GNN Bottleneck
- Graph Equivariance
- Graph Invariance
- Permutation Invariance
- Permutation Equivariance
- Position-Aware GNN
- Identity-Aware GNN
- Higher-Order GNN
- Subgraph GNN
Graph Transformers¶
- Graph Transformer
- Graphormer
- SAN (Spectral Attention Net)
- GPS (General Powerful GNN)
- Laplacian Positional Encoding
- Random Walk Struct Encoding
- Sign-Invariant Network
- Basis-Invariant Network
- Spatial Bias (Graph Attention)
- Centrality Encoding
- Edge Encoding (Transformer)
- Relative Positional Encoding
Knowledge Graphs¶
- Knowledge Graph
- KG Entity
- KG Relation
- KG Triple
- KG Completion
- Link Prediction (KG)
- TransE
- TransR
- DistMult
- ComplEx
- RotatE
- KG Embedding Geometry
- Bilinear KG Model
- Symmetry (Relation Pattern)
- Antisymmetry (Relation)
- Inversion (Relation)
- Composition (Relation)
- Query Embedding
- Box Embedding (Query2Box)
- Multi-Hop Query
- Conjunctive Query
- Neural Bellman-Ford Net
- Inductive KG Reasoning
- ULTRA
- InGram
Heterogeneous Graphs¶
- Heterogeneous GNN
- R-GCN (Relational GCN)
- Heterogeneous Graph Trans
- HAN (Hetero Attention Net)
- Meta-Path
- Relation-Specific Weights
- Basis Decomposition (R-GCN)
- Type-Specific Projection
Applications¶
- Recommender System (Graph)
- Collaborative Filtering
- Matrix Factorization (Rec)
- Neural Collaborative Filter
- LightGCN
- PinSage
- Drug-Drug Interaction
- Protein-Protein Interaction
- Molecular Graph
- Drug Discovery with GNNs
- Social Network Analysis
- Fraud Detection (Graph)
- Traffic Forecasting (Graph)
- Scene Graph
- Relational Database as Graph
- Temporal Graph
- Dynamic Graph
- Temporal GNN (TGN)
- TGAT
- Frequent Subgraph Mining
- Subgraph Isomorphism
- Order Embedding
- SPMiner
Scalability¶
- Mini-Batch Training (GNN)
- Neighbor Sampling
- GraphSAINT
- Cluster-GCN
- LADIES Sampler
- Layer-Wise Sampling
- Historical Embeddings (SIGN)
- Graph Partitioning
Generative Models¶
- Graph Generative Model
- GraphRNN
- GCPN
- Molecule Generation
- Variational Autoencoder (VGAE)
- DiGress
- Graph Generation Metrics
Advanced Topics¶
- In-Context Learning (Graphs)
- Conformalized GNN
- Graph Foundation Model
- LLM + GNN Integration
- Text-Attributed Graph
- Graph Instruction Tuning
- One-For-All (OFA) Model
- Agent Memory as Graph
- Tool-Use Graph
- Relational Deep Learning
- RelBench
- RelGNN
Training & Optimization¶
- Cross-Entropy (Node Class)
- Binary Cross-Entropy (LP)
- Negative Sampling (LP)
- Data Augmentation (Graph)
- Dropout (GNN)
- Batch Normalization (GNN)
- Layer Normalization (GNN)
- DropEdge
- PairNorm
- Early Stopping (GNN)
- Train/Val/Test Split (Trans)
- Train/Val/Test Split (Ind)
- Curriculum Learning (Graph)
Tools & Frameworks¶
- PyTorch Geometric (PyG)
- NetworkX
- Open Graph Benchmark
- SNAP
- DeepSNAP
- DGL
- TUDataset
- PyKEEN
- RelBench (Benchmark)
Self-Supervised & Evaluation¶
- Self-Supervised Learning
- Deep Graph Infomax (DGI)
- Graph Contrastive Learning
- Contrastive Loss
- Graph Augmentation (SSL)
- APPNP
- Simple Graph Conv (SGC)
- Principal Neighborhood Agg
- Mean Reciprocal Rank (MRR)
- Hits@K Metric
- ROC-AUC Score
- Transfer Learning (Graphs)
- Pre-Training (GNN)
- Fine-Tuning (GNN)
- Node Feature Normalization
- Graph Sampling Strategy
- Stochastic Depth (GNN)
- SIGN Architecture