🌍 HealthPredict: AI for Diabetes Risk Prediction (SDG 3)

Repository: ai-se-w02-core-concepts

The project implements a binary classification model (Logistic Regression) to predict the likelihood of Type 2 Diabetes using patient biometric data. This work illustrates key Week 2 learning objectives of the “AI for Software Engineering” specialization (Supervised Learning, Model Evaluation, Ethical Reflection).

---

🎯 Problem & SDG Alignment

Early detection of diabetes plays a vital role in prevention and treatment.
Our AI model acts as a low-cost, high-scale screening tool to identify high-risk patients in resource-limited settings, directly contributing to global health improvement and reducing healthcare system burdens.

🤖 Machine Learning Approach

Component	Detail
Dataset	Pima Indians Diabetes Dataset (diabetes.csv).
Approach	Binary supervised classification (Diabetic / Non-Diabetic).
Model	Logistic Regression (LogisticRegression).
Tech Stack	Python, Pandas, NumPy, Scikit-learn, Matplotlib, Seaborn.

---

📂 Repository Structure

ai-se-w02-core-concepts/
├── src/
│   └── sdg3_health_predictor.py                  # Main Python ML script
├── data/
│   └── diabetes.csv                              # Project dataset (Input)
├── docs/
│   ├── SDG3_Report_Final.md                      # Summary Report
│   └── SDG3_HealthPredict_PitchDeck_2025.pdf     # Final Pitch Deck
├── assets/
│   ├── banner.png                                # Cover
│   └── confusion_matrix.png                      # Model evaluation plot
└── README.md                                     # Repository entry point (This file)

⚙️ Setup and Execution

1. Clone the Repository

git clone https://github.com/software-development-course-2025/ai-se-w02-core-concepts
cd ai-se-w02-core-concepts

2. Move Data File

Place the diabetes.csv file inside the data/ folder.

3. Install Dependencies

pip install pandas numpy scikit-learn matplotlib seaborn

4. Run the ML Script

python src/sdg3_health_predictor.py

The script will print the evaluation metrics and display the Confusion Matrix plot.

📊 Model Results (Unseen Data)

The model was trained on 80% and evaluated on 20% of the data.

Metric	Score (Example Output)
Accuracy	0.7825
Precision	0.7011
Recall	0.6180
F1-Score	0.6572

Confusion Matrix

The visualization below confirms the model's classification performance:

💡 Ethical Reflection and Social Impact

The project highlights:

• Risks of data bias, as the dataset represents a limited population.
• Importance of equity auditing to ensure fair performance across diverse demographic groups.
• Necessity of responsible AI practices, transparency, and careful deployment in clinical environments.

These considerations align with SDG 3 by promoting safe, ethical, and equitable health technology development.

---

ℹ️ Author & License

👤 Author: Augusto Mate
📧 Email: [email protected]
🐙 GitHub
🔗 LinkedIn

This project is licensed under the MIT License.
See the LICENSE file for details.

---

In every dataset lies a story, and in every prediction, a chance to change one.
HealthPredict is a small step toward a future where care begins with foresight.