GNN方法和模型的Pytorch实现。Pytorch implementation of GNN.

A PyG-based package of spectral GNNs with benchmark evaluations (SIGMOD 2026).

This project detects Pneumonia from Chest X-ray images using deep learning. It explores four approaches: A custom CNN built from scratch ResNet with transfer learning Graph Neural Network (GAT) for spatial reasoning A Multimodal model combining global and patch features The dataset used is from Kaggle's Chest X-Ray Pneumonia collection.

Graph-Based Reinforcement Learning for Structural Cross-Sectional Design Optimization

Spatio-Temporal GNN (T-GCN, A3T-GCN) for stock trend classification and price forecasting for portfolio selection.

Implementation in Pytorch of the article "E(n) Equivariant Graph Neural Network" - Satorras et al.

蔡老師的課程作業

Master's Thesis developed at Technische Universität München (TUM) under supervision of Prof. Stephan Günnemann.

Introduction to applied machine learning

Developed a Reinforcement Learning agent for Tetris, achieving a 20% improvement over a heuristic baseline. Engineered state representation and reward shaping, and conducted hyperparameter tuning to improve training stability.