Learn machine learning through hands-on, interactive projects.
Before the models: master the two libraries every ML notebook is built on. Arrays, slicing, broadcasting, and matrix math with numpy; DataFrames, filtering, groupby, and missing-data repair with pandas — no machine learning yet, just every data skill it requires.
Learn the fundamentals of ML through hands-on Python notebooks.
Build linear regression models entirely from scratch — from a single feature to a full multivariate housing-price model — using nothing but numpy and pandas.
Predict whether a tumor is malignant or benign using the classic Breast Cancer Wisconsin dataset — and learn why accuracy alone can mislead, with confusion matrices, precision, recall, and decision thresholds.
Open the box on deep learning: build a small neural network with pure numpy — forward pass, ReLU, backpropagation, gradient descent — and watch it learn the curves a straight line can't, on the same house-price data from the Linear Regression project.
Go from tensors to trained neural networks with PyTorch. Start by letting autograd re-derive the linear-regression gradients you once computed by hand, then build up to real neural nets — running on a dedicated server-side kernel, not in the browser.