Overview
This advanced course presents machine learning through a statistical framework, emphasizing prediction, generalization, interpretability, and responsible model development. Students examine supervised and unsupervised learning methods, including regression, classification, tree-based models, support vector machines, neural networks, clustering, dimensionality reduction, and ensemble methods.
The course develops practical competence in data preprocessing, feature engineering, reproducible computational workflows, model selection, tuning, validation, and introductory deployment and monitoring. Evaluation includes train-validation-test design, cross-validation, bias-variance tradeoffs, calibration, class imbalance, data leakage, performance metrics, uncertainty, and model comparison.
Students use programming and statistical or machine-learning libraries to construct, assess, and communicate models. Ethical and operational considerations include interpretability, fairness, privacy, overfitting, distribution shift, and the responsible communication of limitations and recommendations.
Learning Outcomes
- Translate domain problems into appropriate machine-learning tasks, target variables, and evaluation objectives.
- Prepare reliable datasets by applying suitable methods for cleaning, preprocessing, feature engineering, imbalance management, and leakage prevention.
- Fit and tune linear, logistic, nearest-neighbor, tree-based, boosting, support vector, neural network, clustering, and dimensionality-reduction models.
- Compare and validate competing models using resampling strategies, appropriate performance metrics, calibration analysis, and uncertainty estimates.
- Diagnose overfitting, underfitting, bias-variance tradeoffs, distribution shift, and limitations in model generalization.
- Interpret model behavior using suitable diagnostic, feature-importance, and explanation tools while distinguishing association from causal claims.
- Evaluate fairness, privacy, transparency, and operational risks in machine-learning development and deployment.
- Construct reproducible workflows using programming and statistical or machine-learning libraries, including introductory monitoring procedures.
- Communicate model results, limitations, and recommendations in a technically defensible manner for specialist and non-specialist audiences.
Timetable
| Type | Length | Frequency | Period |
|---|---|---|---|
| Lecture | 2 hours | Weekly | All semester |
| Lab | 2 hours | Weekly | All semester |
| Tutorial | 1 hour | Fortnightly | All semester |
Assessment Schedule
| Type | Description | Weighting |
|---|---|---|
| Quiz | Weekly applied quizzes (6 × 2%) | 12.00% |
| Assignment | Data preparation and exploratory analysis assignment | 15.00% |
| Assignment | Model development and validation assignment | 20.00% |
| Test | Practical programming test | 15.00% |
| Capstone | Machine learning project report and presentation | 18.00% |
| Exam | Final examination | 20.00% |
Prerequisites
Teaching Staff & Programs
This course is delivered jointly by faculty from the participating programs listed below. In line with the Douchewater Way, the University of Sexology tailors core instruction directly to each cohort's specific discipline — adapting curriculum to program needs rather than forcing students into a one-size-fits-all model. Learn more about our approach at The Douchewater Way.
