Data science is the craft of turning raw, messy data into decisions — a blend of statistics, machine learning and computational thinking, aimed squarely at real-world questions. A data scientist cleans and explores a dataset, engineers the right features, fits and validates a model, measures it honestly, and communicates what it means to people who will act on it.

This master's-level course teaches the mathematics and statistics that hold the whole pipeline together — imputation and information theory, cross-validation and ROC curves, boosting and A/B testing, time series and causal inference. It assumes fluency with statistics, the machine learning course, and linear algebra.

We deliberately teach the ideas, not the code. The hands-on Python — pandas, scikit-learn, the whole practical toolkit — is best learned by doing, so throughout the course we point you to the free, interactive notebooks of Kaggle Learn. Understand the maths here; practise the code there. The final module gathers every Kaggle course in one place as a parallel practical track.

Module 1 — Foundations of data science

Practise in Python: Kaggle Learn — Pandas.

Module 2 — Wrangling and cleaning data

Practise in Python: Kaggle Learn — Data Cleaning.

Module 3 — Exploratory data analysis and visualization

Practise in Python: Kaggle Learn — Data Visualization.

Module 4 — Feature engineering and information theory

Practise in Python: Kaggle Learn — Feature Engineering.

Module 5 — Supervised learning and model validation

Practise in Python: Kaggle Learn — Intro to Machine Learning and Intermediate Machine Learning.

Module 6 — Model evaluation and imbalance

Practise in Python: Kaggle Learn — Intermediate Machine Learning.

Module 7 — Ensemble methods and hyperparameter tuning

Practise in Python: Kaggle Learn — Intermediate Machine Learning.

Module 8 — Experiments, inference and causal analysis

Module 9 — Time series analysis and forecasting

Practise in Python: Kaggle Learn — Time Series.

Module 10 — Applied data science

Interpretability, recommenders, ethics, communication and data at scale — the parts that turn a model into a result people trust and use.

Practise in Python: Kaggle Learn — Machine Learning Explainability and Intro to AI Ethics.

Module 11 — Practice on Kaggle

This course teaches the mathematics; Kaggle Learn teaches the code, free and hands-on in the browser. Worked in this order, the two tracks run in parallel — read a Primer module, then do the matching Kaggle course to practise it on real data:

1. Intro to Programming — variables, loops, functions.
2. Python — the language data science runs on.
3. Pandas — loading, filtering and reshaping tables.
4. Data Cleaning — missing values, scaling, dates, inconsistent text.
5. Data Visualization — charts that reveal structure.
6. Intro to SQL and Advanced SQL — querying data at the source.
7. Intro to Machine Learning — your first models and validation.
8. Intermediate Machine Learning — pipelines, cross-validation, XGBoost, leakage.
9. Feature Engineering — building features that make models better.
10. Time Series — trend, seasonality and forecasting.
11. Machine Learning Explainability — permutation importance and SHAP.
12. Intro to AI Ethics — bias, fairness and responsible use.

Begin → What Is Data Science?