Your browser doesn't support speech here. You can still read everything, or try Chrome, Edge or Safari.
Course 4 of 10 · Beginner → Intermediate
Machine Learning
From zero to job-ready ML. Python, NumPy and pandas, the maths you need, core ML algorithms, model evaluation, deployment, LLM basics, tools and libraries, and a job-ready portfolio.
How this course works
Welcome
- Level: Beginner → Intermediate. You need: Python and Maths for AI (or equivalent). SQL helps.
- Starting from zero? Begin with Your Roadmap: Zero to ML Intern. The course goes Core ML → Model quality → Algorithms → Deep learning → Production & LLMs → Job ready → Tools & libraries.
- Before you start: the Python and Maths for AI courses (or pass their checkpoints: Python, Maths). SQL & Databases helps too.
- The story game, First 90 Days at Bluegum Health, runs through the course: each 🎮 milestone is a realistic week as a junior ML engineer.
- Pick a topic from the sidebar; they go in order. Press ▶ Play this topic to hear a lesson read aloud, or 🔊 Explain on one section.
- Topics climb from Beginner to Intermediate · Maths, Advanced and Code; use the Show chips to hide levels you don't need. Most have a Watch it animation, Try it demos and runnable code examples, and every topic ends with a quiz: answer it to mark the topic done ✓.
- Stuck? Press 💬 Ask the AI tutor: a small free AI model that runs on your own device and answers from the course.
- Progress is saved in this browser and shared across all the courses on this site.
Learning path: 🐍 Python → 🗄️ SQL & Databases → ∑ Maths for AI → 📈 Machine Learning → 🧠 Deep Learning → 👁️ Computer Vision → 💬 Natural Language Processing → ✨ Generative AI & LLM Apps → 🚀 Flask & Deployment → 🔬 Becoming an AI Researcher
Getting started
ML foundations
2. What is Machine Learning?Learning patterns from data instead of hand-writing rules.3. Data, Features & SplitsPreparing data properly — where most real ML effort goes.4. Data CleaningFinding and fixing the errors, inconsistencies and gaps in raw data.5. Data Transformation & Feature EngineeringReshaping clean data into the form models learn from best.6. 🎮 Milestone 1: Onboarding & the data auditYour first week at Bluegum Health. Turn a vague request into a real ML problem, then find out what is wrong with the data before anyone trains a model. Includes 9 hands-on Python coding tasks (pandas, scikit-learn), checked by automated tests.7. Linear RegressionThe simplest predictive model — and the foundation for many others.8. Gradient DescentHow almost every ML model, including neural nets, actually learns.9. Logistic RegressionDespite the name, it's a classification model that outputs probabilities.10. scikit-learn EssentialsThe library behind most classical ML: the estimator API, splitting, preprocessing, Pipelines and ColumnTransformers, cross-validation, tuning, metrics and saving models.
Model quality
11. Bias, Variance & OverfittingThe central trade-off in all of machine learning.12. RegularizationPenalising complexity so models generalise better.13. Evaluation Metrics & Cross-ValidationMeasuring performance honestly — accuracy is often not enough.14. 🎮 Milestone 2: Baseline, splits & the right metricSplit the data the way it will be used, build your first models (a baseline, logistic regression, then a proper preprocessing pipeline), forecast with linear regression and gradient descent, regularise with ridge and lasso, fix a broken k-NN, and only then evaluate: agree the metric with the clinic, plot the results and choose a cost-based threshold. Ends with communication, code review and debugging. Includes 7 hands-on Python coding tasks (pandas, scikit-learn), checked by automated tests.15. ✅ Checkpoint 3: Core machine learningA mixed quiz on the whole core machine learning phase. Pass mark: 6 of 9 on the first try.
Supervised algorithms
16. k-Nearest NeighboursClassify a point by asking its closest neighbours to vote.17. Decision TreesA flowchart of yes/no questions learned from data.18. Ensembles: Random Forests & BoostingMany weak or noisy models combined into one strong model.19. Support Vector MachinesFind the boundary with the widest possible margin.
Unsupervised learning
20. Clustering & k-MeansFinding natural groups in unlabelled data.21. Dimensionality Reduction & PCACompressing many features into a few that keep most of the information.22. 🎮 Milestone 3: Improving the model, fairlyRun a fair bake-off between decision trees, random forests, gradient boosting and SVMs, tune without fooling yourself, review fairness, segment doctors with PCA and k-means, and write the model card. Includes 7 hands-on Python coding tasks (pandas, scikit-learn), checked by automated tests.
Applied ML
23. Time Series ForecastingTrend, seasonality, baselines, lag features, time-based validation and backtesting: predicting demand, visits and sales.24. Recommender SystemsPopularity baselines, content-based and collaborative filtering, matrix factorisation, implicit feedback, cold start and ranking metrics.25. Anomaly DetectionFinding unusual cases with few or no labels: robust z-scores, isolation forests, density and distance methods, and human review.26. ✅ Checkpoint 4: AlgorithmsA mixed quiz on the whole algorithms phase. Pass mark: 6 of 8 on the first try.
Deep learning
27. Neural NetworksLayers of simple units that together learn complex functions.28. Training Neural NetworksBackpropagation, activations, optimisers and the tricks that make deep nets work.29. CNNs, RNNs & TransformersSpecialised architectures for images, sequences and language.30. Deep Learning in Practice (PyTorch)Tensors, autograd and the training loop; data loading, GPUs, transfer learning and fine-tuning, with three Colab notebooks.31. 🎮 Milestone 4: A deep-learning project, safelyTriage patient messages with a transformer and route scanned documents with a CNN: choose models against real constraints, label data, diagnose training, and never miss an urgent message. Includes 4 hands-on Python coding tasks (pandas, scikit-learn), checked by automated tests.32. ✅ Checkpoint 5: Deep learningA mixed quiz on the whole deep learning phase. Pass mark: 6 of 8 on the first try.
Beyond the basics
33. Reinforcement LearningLearning by trial, error and reward.34. The ML Project WorkflowPutting it all together, from problem to deployed model.35. Deploying Models: APIs, Docker, Tracking & CIFrom a notebook to a service: saving models, batch vs real-time, FastAPI, input validation, Docker, experiment tracking, CI and monitoring.36. 🎮 Milestone 5: Launch, A/B test & on-callLaunch with an A/B test, let a reinforcement-learning bandit choose reminders safely, handle a 2 a.m. production incident, and write a blameless postmortem. Includes 5 hands-on Python coding tasks (pandas, scikit-learn), checked by automated tests.37. LLMs, Embeddings & RAGHow large language models work, how embeddings power semantic search, and how to build, evaluate and secure a retrieval-augmented assistant.38. 🎮 Milestone 6: An LLM assistant, safelyRiverside's front desk is drowning in policy questions. Choose between RAG, fine-tuning and prompting, build retrieval and a grounded prompt in Python, defend against prompt injection, and evaluate the assistant before anyone relies on it. Includes 3 hands-on Python coding tasks (pandas, scikit-learn), checked by automated tests.
Job ready
39. Capstone Project & Interview PrepTurn what you have learned into a portfolio project that proves it, and prepare for each stage of an ML engineering interview.40. Portfolio Projects: Build, Publish, PresentTwo guided end-to-end projects on real public data, a repository template, a README that gets read, and how to present your work.41. Python Coding Interview PracticeHow coding interviews work, a method that works under pressure, Big-O basics, the common patterns, and 20 graded problems in the browser.42. 🎮 Milestone 7: Interview dayYour contract converts to a permanent role, but only through Bluegum's full interview loop. A recruiter screen, rapid-fire ML questions, a timed coding round, an ML system design interview and a behavioural round, using the work you have done in this game. Includes 1 hands-on Python coding task (pandas, scikit-learn), checked by automated tests.43. ✅ Checkpoint Intern-readinessA mixed quiz on the whole intern-readiness phase. Pass mark: 8 of 10 on the first try.
Tools & libraries
44. Tools & Libraries: Start HereA short index: which package does which job and where it is taught. NumPy, pandas, plotting, SQL and scikit-learn stay in their own topics; this section adds PyTorch, Hugging Face and gradient boosting.45. PyTorch EssentialsThe library step by step: tensors and devices, autograd, nn.Module, Dataset and DataLoader, the training loop, evaluation mode, saving and loading, and the errors everyone hits.46. Hugging Face: Pretrained Models in PracticeThe Hub, pipelines, tokenizers, Auto classes, the datasets library, fine-tuning with Trainer, LoRA with peft, and reading model cards and licences.47. Gradient Boosting Libraries: XGBoost & LightGBMWhy boosted trees usually win on table data, how XGBoost, LightGBM and scikit-learn compare, the parameters that matter, early stopping, categorical features, and explaining the model.
When you finish
Next course
🧠 Deep LearningBuild, train and debug neural networks. Tensors, autograd and networks from scratch; optimisers, regularisation and debugging; CNNs, RNNs and transformers; transfer learning, generative models and efficient deployment.🚀 Flask & DeploymentPut your models online. How the web works, Flask apps, templates and forms, databases, REST APIs, serving an ML model, testing, Docker and deploying to the cloud with monitoring.