PyG Heterogeneous Graph Practice

Homogeneous Graph and Heterogeneous Graph

Homogeneous graph

Without distinguishing between node and edge types, there is only one type of node and one type of edge.

Node Type + Edge Type = 2

For example, in a social network, it can be imagined that nodes only have one category 'person', and edges only have one type of connection 'knows'. And people either know each other or they do not.

PyTorch Practical for Linear Neural Network

• 1. Implementing Linear Regression with PyTorch
• 2. Implementing Softmax Regression with PyTorch

Deep Learning Practice with PyTorch

• 0. PyTorch's APIs
• 1. Data Loading and Preprocessing
• 2. Defining Network Models
• 3. Defining Loss Function and Optimizer
• 4. Training the Network
• 5. Testing the Network
• 6. Saving and Loading Models
• 7. GPU Acceleration
• 8. Visualization with TensorBoard

Tensor Operations

• 1. Scalar, Vector, Matrix, and Tensor
• 2. Initializing Tensors
• 3. Attributes of Tensors
• 4. Basic Operations on Tensors
• 5. Summation and Averaging
• 6. Product Operations
• 7. Calculating the Norm of a Vector
• 8. Gradient Computation

Neural Network Training Key Points Interpretation

• 1. Overall Objective
• 2. Loss Function(Objective Function): Quantifying Model Effectiveness
• 3. Optimization Algorithms (Gradient Descent): Algorithms that Adjust Model Parameters to Optimize the Objective Function
• 4. Hyperparameters

Artificial Intelligence Concept Interpretation

• 1. Machine Learning, Deep Learning, and Reinforcement Learning
• 2. Supervised Learning, Semi-Supervised Learning, Unsupervised Learning
• 3. Online Learning and Offline Learning
• 4. Regression, Classification, and Multi-label Classification
• 5. Recommendation and Search