Python smuggled into Andrew Ng’s MATLAB-era ML course

What it does

The repository translates the programming assignments from Andrew Ng’s popular Coursera Machine Learning MOOC into Python-based Jupyter Notebooks. Each exercise lives in its own folder as an .ipynb file paired with a utils.py helper, and the original MATLAB/OCTAVE instructions have been rewritten to reference NumPy, SciPy, and Matplotlib instead. Crucially, the notebooks still talk to Coursera’s original autograder, so you can submit parts for grading without touching the official MATLAB materials.

The interesting bit

The project is essentially a careful translation layer: it preserves the pedagogical structure of a course built before Python dominated ML, while swapping out the entire toolchain. The submission code is partly derived from an earlier community Python port, but the integration with Coursera’s live grader means these notebooks are drop-in replacements, not just study guides.

Key highlights

• Re-implements all programming assignments from the MOOC as self-contained Jupyter Notebooks.
• Embedded instructions replace every MATLAB/OCTAVE reference with its Python equivalent.
• Notebooks submit directly to Coursera’s original autograder, so the Python work gets graded like the MATLAB version.
• Each folder pairs an .ipynb notebook with a utils.py helper module; you only edit the notebook.
• Deepnote integration lets you run assignments in a browser without local setup.

Caveats

• The README only shows folder names like Exercise1 and never spells out how many assignments are included or what topics they cover.
• The stated requirements freeze a 2018-era Python stack (3.6, NumPy 1.13, etc.); later versions are suggested but not explicitly tested in the text.
• Part of the submission code relies on prior work by mstampfer, so this is a polished community translation rather than a from-scratch grader implementation.

Verdict

Grab this if you are taking Andrew Ng’s MOOC and would rather write Python than OCTAVE. Skip it if you are looking for original coursework or a general-purpose ML tutorial unrelated to the Coursera class.