tangyudi/Ai-Learn

The curriculum structures learning through a pedagogical progression from foundational skills (Python, linear algebra, probability) through classical machine learning algorithms to modern deep learning frameworks (TensorFlow 2, PyTorch, Keras), with practical implementation emphasizing feature engineering and hyperparameter tuning over theoretical derivation. The repository bundles ~200 hands-on case studies spanning tabular data mining, computer vision (OpenCV-based image processing and CNN applications), and NLP tasks, distributed via Jupyter notebooks with accompanying datasets and presentation materials designed around real-world competition solutions and employment interview preparation. Integration focuses on scikit-learn for traditional ML workflows, Pandas/NumPy for data manipulation, and production-grade frameworks for deep learning, with particular emphasis on gradient boosting libraries (XGBoost, LightGBM) for tabular data competitions.

12,749 stars. No commits in the last 6 months.