Python for Machine Learning · PyLernen
PyLernen logo Python for Machine LearningPyLernen · Data, models, deployment & responsible AI
Machine Learning
2 of 14 modules

Self-paced path · 14 modules · 5 workspaces · 10 applied projects

Python for Machine Learning

Go from Python basics to models you can deploy and defend. Every lesson has code you can edit and run right here, exercises that check your answer, and a plain-language explanation of what the output means.

Each node is a module. It lights up when you finish it.2 of 14 complete
Continue with Module 3Open a workspace

How every lesson works

  1. ConceptThe idea in plain words.
  2. ExplanationHow it works and why.
  3. CodeA complete example you can edit.
  4. RunPress Run and read the output.
  5. InterpretWhat the numbers mean.
  6. ModifyChange something and predict the effect.
  7. ExerciseWrite code; it is checked for you.
  8. ChallengeA harder, open task.
  9. Knowledge checkA short quiz per module.

How the Run button works

Simulator

By default, code runs in PyLite, a Python simulator built into this page. It really parses and executes your code line by line, so if you change a number, a dataset or a setting, the output changes. It supports the parts of Python, NumPy, pandas, scikit-learn, Matplotlib and Seaborn used in this course, plus small PyTorch and Keras subsets.

Matches Python

Most outputs are identical to real Python, because the simulator follows the same algorithms and the same random-number generators (for example train_test_split(..., random_state=42) picks the same rows).

Approximate

Some results cannot be reproduced exactly in a browser, such as neural-network training or optimisers that stop early. When that happens the output panel says so, and where possible shows the output recorded from real Python next to it.

Real Python

Each code box also has a Real Python (Pyodide) option. It downloads the real Python interpreter into your browser. If this page’s security settings block that download, the box tells you and keeps using the simulator.

Modules

01Python Foundations for ML4 lessons · 4 exercises (4 passed)Complete02Math & Statistics for ML4 lessons · 4 exercises (4 passed)Complete03Data Analysis with NumPy, pandas, Matplotlib & Seaborn4 lessons · 4 exercises (2 passed)≈ 5 hours04Understanding Machine Learning4 lessons · 4 exercises (0 passed)≈ 3 hours05Supervised Learning: Regression4 lessons · 4 exercises (0 passed)≈ 5 hours06Supervised Learning: Classification5 lessons · 4 exercises (0 passed)≈ 6 hours07Unsupervised Learning4 lessons · 3 exercises (0 passed)≈ 4 hours08Model Evaluation & Optimisation4 lessons · 4 exercises (0 passed)≈ 5 hours09Advanced Machine Learning4 lessons · 4 exercises (0 passed)≈ 5 hours10Deep Learning4 lessons · 3 exercises (0 passed)≈ 6 hours11Applied ML Projects2 lessons · 1 exercises (0 passed)≈ 2 hours + projects12Deploying Models3 lessons · 2 exercises (0 passed)≈ 3 hours13Responsible AI & Data Privacy4 lessons · 2 exercises (0 passed)≈ 3 hours14Monitoring & MLOps4 lessons · 2 exercises (0 passed)≈ 3 hours

Workspaces

Hands-on benches where you change settings and immediately see what a model does. Each one also shows the Python that produces the result, so you can run it and change it.

WorkspaceRegression LabFit lines and curves, watch gradient descent find the slope.WorkspaceClassification LabTrain classifiers and see their decision boundaries.WorkspaceClustering LabWatch k-means move its centres; compare with DBSCAN.WorkspaceData Cleaning LabFix a messy table step by step with real pandas code.WorkspaceNeural Network LabBuild a small network and watch it learn layer by layer.

Applied projects

P1 · RegressionHouse Price PredictionEstimate a fair asking price from size, rooms, age, location and distance.BeginnerP2 · Regression + classificationStudent PerformancePredict final scores and identify students who need support early.BeginnerP3 · ClassificationCustomer Churn PredictionFind customers likely to leave so the retention team can call them first.IntermediateP4 · Text classificationSpam DetectionFilter junk messages without losing real ones.BeginnerP5 · ClusteringCustomer SegmentationGroup shoppers into segments marketing can act on.IntermediateP6 · Imbalanced classificationFraud DetectionCatch rare fraudulent card payments without drowning investigators.IntermediateP7 · Time-series regressionSales ForecastingForecast next month's unit sales from trend, season, price and advertising.IntermediateP8 · Computer visionImage ClassificationTeach a model to recognise circles, crosses and bars in tiny 8×8 images.IntermediateP9 · Text classificationSentiment AnalysisTell positive product reviews from negative ones automatically.BeginnerP10 · Full workflowEnd-to-End ML PipelineFrom raw data to a tuned, saved, documented and monitored model.AdvancedCompleted
After Module 14

Final exam

30 questions. 70% to pass.

Your reward

Certificate of completion

Unlocks when every module is complete and the exam is passed.