Applied workflow
Students bring together question definition, predictor timing, model fitting, validation and reporting.
Module 5
Applied Biostatistical ML Case Studies
This module applies the full course workflow to realistic medical and biomedical machine learning problems: clinical risk prediction, survival outcomes, high-dimensional omics, missing data, imbalance, fairness and final project reporting.
Module aim
Turn modelling knowledge into complete applied reports.
The purpose of this module is to help students move from learning methods to completing transparent, clinically sensible and statistically responsible machine learning analyses.
5
Lessons
R
Applied project
Preparing
Module status
Case studies
Core focus
Health-data complexity
The module introduces practical issues such as censoring, omics features, missing data, imbalance and subgroup performance.
Report-ready thinking
Each case study is framed around interpretation, limitations and responsible conclusions rather than model output alone.
Module lessons
Apply the full course to realistic health-data problems.
Each lesson turns the earlier modules into applied thinking: defining the clinical question, handling complex data structures, validating honestly, interpreting carefully and reporting with limitations.
5.1
Lesson 5.1
Clinical risk prediction case study
Apply the full prediction workflow to a clinical risk prediction problem, from question definition to validation and reporting.
Open lesson →
5.2
Lesson 5.2
Survival prediction and censored outcomes
Study how prediction changes when outcomes are time-to-event, censored and linked to follow-up time rather than simple binary labels.
Open lesson →
5.3
Lesson 5.3
High-dimensional omics and feature selection
Understand prediction problems with many biomarkers, genes or molecular features, and why feature selection must be handled carefully.
Open lesson →
5.4
Lesson 5.4
Missing data, imbalance and fairness
Learn how missingness, class imbalance and subgroup performance affect responsible medical machine learning.
Open lesson →
5.5
Lesson 5.5
Final applied ML project in R
Bring the course together through a final applied project with data preparation, modelling, validation, interpretation and reporting.
Open lesson →
Final route
Complete the course through applied reporting.
This final module is designed to connect every previous idea: supervised learning, validation, calibration, modern model comparison and responsible interpretation in real health-data settings.
Open final project →Course pathway
Return to the full ML in Biostatistics course.
Use the course homepage to review all five modules, open the case studies, download course assets and return to earlier lessons.
Back to course →