Machine Learning Algorithms from Scratch
Chinland InfoTech Academy flagship — build ML algorithms with NumPy only, from linear regression to decision trees, MLPs, K-Means, and PCA. 45 lessons, 360+ drills with full math derivations.
Who it's for
Learners who want to understand what happens inside fit() and predict() — not just call sklearn.
What you'll learn
- Implement fit and predict in NumPy for regression, classification, trees, neural nets, and clustering
- Derive and code gradients, loss functions, and update rules from first principles
- Build full pipelines: data → preprocess → model → predict → evaluate on academy and shop data
- Explain why each step exists and compare scratch implementations to scikit-learn references
- Deliver academy and shop capstone projects without black-box APIs
Course contents
- 01Foundations · 6 lessons
- 02Linear Regression · 7 lessons
- 03Regularization & Features · 3 lessons
- 04Classification · 10 lessons
- 05Probabilistic & Tree Models · 6 lessons
- 06Neural Networks · 4 lessons
- 07Unsupervised from Scratch · 4 lessons
- 08Integration & Capstone · 5 lessons