Skip to content

HealthPredict Project v1.0.0: Initial Release

Latest

Choose a tag to compare

@augusto-mate augusto-mate released this 16 Nov 14:56
b1293f5

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.