This release marks the official launch of HealthPredict, a proof-of-concept AI solution aligned with UN SDG 3: Good Health and Well-being, focused on early prediction of Type 2 Diabetes risk.
✨ Key Features & Deliverables
1. Core Machine Learning Model:
-
Algorithm: Logistic Regression (Supervised Classification)
-
Functionality: Predicts diabetes risk (binary outcome) based on 8 key clinical/biometric features from the Pima Indians Diabetes dataset
-
Code: Fully commented Python script
src/sdg3_health_predictor.py(including data preprocessing, training, and robust evaluation)
2. Comprehensive Documentation:
-
README.md: complete repository overview, setup, project structure, results, visuals, ethics
-
Project Report:
docs/SDG3_Report_Final.md
→ 1-page summary (approach, metrics, SDG alignment, ethical reflection) -
Pitch Deck:
docs/SDG3_HealthPredict_PitchDeck_2025.pdf
→ concise storytelling for stakeholders
Performance Metrics
- Accuracy: 0.7825
- Precision: 0.7011
- Recall: 0.6180
- F1-Score: 0.6572
3. Ethical Reflection:
- Focus: Explicit discussion on the risks of data bias and the mandate for continuous auditing and retraining with ethnically diverse datasets to ensure equitable deployment and promotion of global health justice.
🚀 Getting Started
Instructions for installation, dependencies, and model execution are provided in the Setup and Execution section of the README.md.