Level 2 of 3
Intermediate Level
Prerequisites
Machine Learning Beginner level, or equivalent: pandas, train/test splits, basic scikit-learn models and classification metrics.
Duration
8 weeks · 90 guided hours · 6 modules. Taught live by an instructor — online or in person.
By the end of this level you can
- Build leak-free scikit-learn pipelines with feature engineering
- Compare models fairly with cross-validation and appropriate metrics
- Tune hyperparameters without overfitting the validation data
- Find structure in unlabelled data with clustering and dimensionality reduction
- Handle imbalanced, missing and shifting real-world data
Modules
- 1Module 1 · 16 h · 3 lessonsFeature Engineering and Pipelines
- 2Module 2 · 15 h · 3 lessonsCross-Validation and Model Selection
- 3Module 3 · 13 h · 3 lessonsHyperparameter Tuning
- 4Module 4 · 15 h · 3 lessonsUnsupervised Learning
- 5Module 5 · 14 h · 3 lessonsImbalanced and Messy Real-World Data
- 6Module 6 · 17 h · 3 lessonsIntermediate Machine Learning Project
Intermediate level project
Mobile-Money Agent Liquidity Forecast and Segmentation (training data)
Using a provided synthetic data set of mobile-money agents, segment agents by behaviour with clustering and build a tuned model that predicts which agents will run short of cash next week.
Deliverables
- A scikit-learn pipeline with preprocessing and model, evaluated with cross-validation
- A tuned final model with a documented search and a held-out test result
- An agent segmentation with named, interpreted clusters
- A short report with recommendations, limits and a monitoring plan
Assessed on
- All preprocessing is inside pipelines; no leakage between folds
- Model choice and threshold are justified by the cost of each error
- Clusters are interpreted in business terms and checked for stability
- The test set is used once, at the end
All levels are taught live by an instructor — online or in person — see registration options.