Articles
Long-form, code-first writing on machine learning — from gradient descent from scratch to shipping RAG systems in production.
How to Load Machine Learning Data From Scratch In Python
Before you train a model, you have to load data. Read CSVs, convert types and encode labels with nothing but the standard library. Runnable in your browser.
How to Scale Machine Learning Data From Scratch With Python
Many ML algorithms expect data on a common ruler. Implement normalization and standardization from scratch, avoid leakage, and know which one each algorithm wants. Runnable in your browser.
How to Implement Resampling Methods From Scratch In Python
Estimate how your model will perform on unseen data. Implement train/test splits, k-fold, stratified folds and the bootstrap from scratch. Runnable in your browser.
How To Implement Machine Learning Metrics From Scratch in Python
Once you make predictions, you need to know if they are any good. Implement accuracy, the confusion matrix, MAE, RMSE and R² from scratch — and learn which to trust when. Runnable in your browser.
How To Create an Algorithm Test Harness From Scratch With Python
We cannot know which algorithm will be best for a given problem. Build a reusable train/test and cross-validation harness that compares algorithms fairly. Runnable in your browser.
How to Implement Linear Regression From Scratch in Python
The core of many ML algorithms is optimization. Implement stochastic gradient descent from scratch and use it to fit a linear regression model.