Level 3 of 3
Advanced Level
Prerequisites
Machine Learning Intermediate level, or equivalent: pipelines, cross-validation, tuning and clustering with scikit-learn. Basic command-line and Git skills are helpful.
Duration
12 weeks · 140 guided hours · 6 modules. Taught live by an instructor — online or in person.
By the end of this level you can
- Build and interpret high-performing ensemble models
- Explain models and audit them for fairness, and document them responsibly
- Forecast time series with appropriate validation
- Deploy models as reproducible services
- Track, monitor and govern models in production
Modules
- 1Module 1 · 22 h · 4 lessonsEnsemble Methods
- 2Module 2 · 22 h · 4 lessonsInterpretability, Fairness and Documentation
- 3Module 3 · 22 h · 4 lessonsTime Series and Forecasting
- 4Module 4 · 24 h · 4 lessonsDeploying Machine Learning Models
- 5Module 5 · 20 h · 4 lessonsMLOps and Monitoring
- 6Module 6 · 30 h · 5 lessonsMachine Learning Capstone
Advanced level project
Market Price Early-Warning Service
A deployed forecasting service that predicts next month's staple food prices for several markets from a provided training data set, explains its forecasts, is monitored for drift and is documented with a model card.
Deliverables
- Forecasting models validated with time-based backtesting against baselines
- A FastAPI service serving forecasts from a versioned model, packaged to run on another machine
- Experiment tracking records, a monitoring report and a retraining procedure
- A model card covering intended use, performance, limits and fairness checks
Assessed on
- Forecasts beat seasonal-naive baselines in backtests, reported honestly per market
- The service runs from the README on a clean machine and validates its inputs
- Monitoring detects a simulated drift scenario
- The model card is accurate and readable by a non-specialist
All levels are taught live by an instructor — online or in person — see registration options.