The 16-week curriculum
Every week opens on its first day. Your place is remembered in this browser, so you can stop and come back.
Start here
Pick any week. Week 1 assumes no prior experience.
Foundations
Python, data wrangling and the maths underneathPython for Machine Learning
Set up the environment, learn the array and the DataFrame, and build a working churn model before the week is out.
7 daysStart day 1 → Week 02Not startedData Wrangling and Exploratory Analysis
Load, reshape, clean and profile data properly, the work that decides whether a model has anything to learn from.
7 daysStart day 1 → Week 03Not startedThe Maths That Actually Matters
Linear algebra, calculus, probability and statistics, only the parts that change what you do, each one implemented rather than proved.
7 daysStart day 1 →Supervised learning
Regression, classification, trees and ensemblesRegression, Regularisation and Leakage
Predict a number, measure it honestly, and meet the two failures that make a good-looking model worthless.
7 daysStart day 1 → Week 05Not startedClassification Fundamentals
Logistic regression, kNN, naive Bayes and SVMs, and the threshold decision that matters more than any of them.
7 daysStart day 1 → Week 06Not startedTrees, Forests and Boosting
Decision trees, bagging, random forests and gradient boosting, and an honest look at when they lose to a straight line.
7 daysStart day 1 →Evaluation and features
Validation you can trust, and features that beat modelsModel Evaluation and Validation
Measuring a model honestly, the folds, the leaks, the imbalance, the calibration and the segments the average score hides.
7 daysStart day 1 → Week 08Not startedFeature Engineering
Turning raw columns into inputs a model can use, and measuring honestly enough to know which ones helped.
7 daysStart day 1 →Unsupervised
Clustering, dimensionality reduction and anomaliesUnsupervised Learning and Clustering
Finding structure without labels, k-means, hierarchical, DBSCAN and mixtures, plus how to tell whether the structure is real.
7 daysStart day 1 → Week 10Not startedDimensionality Reduction and Anomaly Detection
Compressing wide data without losing what matters, and finding the rows that do not belong.
7 daysStart day 1 →Deep learning
Neural networks, convolution, text and sequencesNeural Networks From Scratch, Then Keras
Build a network in numpy, derive backpropagation, then do it properly in TensorFlow, and find out where it wins.
7 daysStart day 1 → Week 12Not startedConvolutional Networks and Representation Learning
Build convolution from scratch, train a real image classifier, transfer a pretrained base, and settle when deep learning is worth it.
7 daysStart day 1 → Week 13Not startedText, Embeddings and Sequence Models
Turn text into numbers, discover what a bag of words cannot express, and find out whether a sequence model is worth it.
7 daysStart day 1 →Production
Tuning, interpretability, deployment and monitoringHyperparameter Optimisation and Interpretability
Spend a tuning budget properly, then explain what the tuned model is doing, and where explanation stops.
7 daysStart day 1 → Week 15Not startedDeployment and MLOps
Package a model, serve it over HTTP, validate what arrives, watch it drift and replace it safely.
7 daysStart day 1 →Capstone
Your own project, and getting hiredEvery week ends with a self-assessment quiz, and each day can be marked complete as you go. Nothing is sent anywhere; your progress lives in this browser only.
Programme detailsAsk a question