ML boot camp: 100 days of copy-paste or actual learning?

What it does This repo is a day-by-day Jupyter Notebook curriculum covering machine learning from Python basics through regression, classification, clustering, NLP, deep learning, and feature engineering. Each “day” is a self-contained notebook mixing explanations with runnable code. The structure suggests a classroom or cohort-based course rather than a reference manual.

The interesting bit The project leans hard on progressive disclosure — it introduces sklearn pipelines and statsmodels before explaining the linear algebra underneath. That either respects the “just make it work first” pedagogy or papers over fundamentals, depending on your theology. The notebooks include handwritten-style markdown explanations alongside code, which feels more like a teacher’s lecture notes than polished documentation.

Key highlights

• Covers the full scikit-learn ecosystem end-to-end in one linear path
• Includes practical topics often skipped in courses: EDA, feature engineering, encoding, scaling
• Deep learning section touches CNNs and RNNs without requiring PyTorch/TensorFlow fluency first
• 2,700+ stars suggest it actually gets used by students rather than just starred and abandoned
• Appears designed for a specific course cohort (the “CampusX” branding), which may explain the pacing

Caveats

• No description, topics, or README detail provided in sources; actual depth and quality of explanations is unclear from metadata alone
• “100 days” is a marketing frame — actual time commitment and rigor per topic is unspecified
• Jupyter-only format means no tests, no package structure, no reproducible environment specification visible

Verdict Worth a look if you’re a beginner who needs a forced march through ML with guardrails, or an instructor seeking notebook templates to adapt. Skip if you want mathematical foundations first, or need production-grade code patterns rather than educational snippets.