DSI-AFRICA-MUDSReH Health Datathon 2026
AI for Africa: Evaluating Artificial Intelligence Through an African Lens
6th DS-I Datathon Registration
Scan the QR code below to register for the 6th DSI-AFRICA-MUDSReH Health Datathon
Scan this QR code to access the registration form
Africa is not a testing ground — it is a launchpad.
The Data Science Initiative (DSI) and the Mbarara Data Science Research Hub (MUDSReH) are pleased to present the DSI-Africa-MUDSReH Health Datathon, a landmark collaborative event that convenes leading talent across Africa in data science, medicine, engineering, and public health.
This initiative is dedicated to addressing one of the continent’s most critical challenges: the evaluation and contextual adaptation of Artificial Intelligence (AI) models for deployment in real-world African healthcare settings.
This is not just a competition, it is a shared mission to question, test, and improve AI tools so they reflect African patient realities and support clinicians and health systems with solutions that are relevant, inclusive, and impactful.
Purpose & Motivation
Why This Datathon? Why Now?
AI in healthcare is advancing rapidly, especially for disease detection and classification. Yet many foundational models are trained and validated on datasets that do not sufficiently represent African populations.
This underrepresentation creates risks in validity, generalisability, and equitable application of AI within African healthcare systems.
The Aim of the Datathon
The datathon will produce validated AI models tailored to African healthcare contexts through rigorous evaluation of existing models using representative African datasets. Teams will identify performance gaps, biases, and limitations, then adapt and fine-tune models to improve local relevance, accuracy, and reliability.
Who Should Join?
Academia
Researchers, PhD and Masters candidates, lecturers.
Clinical Practice
Medical doctors, nurses, radiologists, public health officers.
Technology
Data scientists, ML engineers, biostatisticians, developers.
DSI-AFRICA-MUDSReH Datathon Program 2026
AI for Africa: Evaluating Artificial Intelligence Through an African Lens
16th - 17th May 2026 • Barracuda Foyer
Event Details
DAY 1: Training, Orientation & Team Formation
Barracuda Foyer • 9:00 AM – 1:00 PM (morning sessions)Opening & Welcome
Introduction to the datathon, facilitators, objectives, and expected outcomes.
Why Datathons Matter in Healthcare
Role of datathons in innovation, rapid prototyping, collaboration, healthcare impact, and translating data into actionable insights.
Introduction to Health Data Science Workflow
End-to-end pipeline: problem definition → data collection → EDA → feature engineering → modeling → deployment → interpretation.
Exploratory Data Analysis (EDA) for Health Data
Understanding variables, missing data, class imbalance, outliers, distributions, visualization, correlations, and generating insights. Includes quick demos using simple healthcare datasets.
Team Formation
Group formation and assigning initial team roles.
Tea Break ☕
Networking and informal discussions.
Expectations & Rules of Engagement
Event structure, judging criteria, deliverables, timelines, ethics, teamwork expectations, and communication channels.
Building Simple ML Models
Introduction to baseline ML models (Logistic Regression, Decision Trees, Random Forests). Importance of validation, overfitting, and evaluation metrics in healthcare.
Vector Embeddings & Modern Health AI
Intuitive introduction to embeddings, representation learning, text/image embeddings, retrieval systems, and their applications in healthcare AI.
Computing Infrastructure & Deployment
GPUs vs CPUs, cloud vs local compute, notebooks, APIs, model serving, lightweight deployment strategies, and practical considerations during the datathon.
Dataset Walkthrough & Challenge Description
Description of datasets, target tasks, expected outputs, constraints, ethical considerations, and available resources.
Team Formation & Project Ideation
Group formation, brainstorming project ideas, assigning roles (EDA, modeling, deployment, presentation, etc.).
Final Instructions & Kickoff
Submission guidelines, checkpoints, mentoring structure, and official start of the datathon project work.
Afternoon & Evening
Team Project Work Begins
Participants work independently in teams with mentor support available periodically.
Suggested milestones:
- Data cleaning & EDA
- Baseline model
- Advanced improvements
- Deployment/demo prototype
- Presentation preparation
DAY 2: Evaluation, Fine-tuning, and Presentations
Barracuda FoyerHands-on Work Session II
Continue model evaluation and adaptation with mentor guidance.
Break
Validation and Fairness Checks
Teams prepare benchmark metrics, error analysis, and context-specific interpretation.
Lunch
Finalization of Team Outputs
Teams prepare presentation decks and submission artifacts.
Team Presentations and Feedback
Jury evaluation and panel feedback.
Awards, Reflections, and Way Forward
Closing reflections and next steps for model deployment research.
Closing
Dinner
Networking and team check-ins.
Important Notes
- All participants should come with laptops and charger/accessories
- Teams are multidisciplinary across health, data science, and engineering
- Focus is on evaluating and adapting AI models using representative African datasets
- Mentors and clinicians will support technical and domain-specific decision-making
Share Your Feedback
We value your input! After participating, please complete the 2026 evaluation form.
Submit Evaluation