🚀 Diabetes Risk Prediction System

🌐 Live Demo

👉 Deployed App:
https://diabetes-prediction-system-amwusbsaqrcspfgv6zskka.streamlit.app

---

📌 Project Overview

The Diabetes Risk Prediction System is an end-to-end Machine Learning application that predicts the likelihood of diabetes using patient medical attributes.

This project demonstrates:

• Data preprocessing & cleaning
• Feature scaling
• Model comparison
• Hyperparameter tuning
• Model evaluation (ROC, AUC, Confusion Matrix)
• Model deployment using Streamlit Cloud

This is a production-style ML workflow, not just a notebook experiment.

---

🧠 Dataset

• Pima Indians Diabetes Dataset
• 768 samples
• 8 medical predictor features
• Binary classification:
  • 0 → No Diabetes
  • 1 → Diabetes

---

🔎 Models Implemented

Model	Accuracy	ROC-AUC
Logistic Regression	71%	0.66
Random Forest (Tuned)	75%	0.74

Random Forest was selected as the final deployed model after hyperparameter tuning.

---

⚙️ Machine Learning Pipeline

1. Data Cleaning (zero-value replacement with median)
2. Feature Scaling using StandardScaler
3. Train-Test Split
4. Baseline Model: Logistic Regression
5. Random Forest with:
   • class_weight='balanced'
   • GridSearchCV for tuning
6. Cross-validation
7. Model Evaluation:
   • Accuracy
   • Precision
   • Recall
   • F1-score
   • ROC-AUC
8. Model Saving using joblib
9. Deployment using Streamlit Cloud

---

📈 Model Evaluation

Final Random Forest Performance:

• Accuracy: 74.6%
• ROC-AUC: 0.745
• Balanced performance on imbalanced dataset
• Confusion Matrix analysis
• False Positive & False Negative evaluation

The deployed application also includes:

• ROC Curve visualization
• Feature Importance chart
• Probability confidence scores

---

💻 Web Application Features

• Interactive patient input form
• Model selection (Logistic Regression / Random Forest)
• Risk prediction output (Low / High Risk)
• Confidence probability display
• ROC Curve visualization
• Feature importance chart
• Medical disclaimer
• Clean production UI (Dark theme)

---

🛠 Tech Stack

• Python
• Pandas
• NumPy
• Scikit-learn
• Matplotlib
• Seaborn
• Joblib
• Streamlit
• Git & GitHub
• Streamlit Cloud

---

📂 Project Structure

Diabetes-Prediction-System/ │ ├── Data/ │ └── pima-indians-diabetes.csv │ ├── app.py ├── Train_Model.py ├── diabetes_model.joblib ├── logistic_model.joblib ├── scaler.joblib ├── requirements.txt └── README.md

⚠ Medical Disclaimer

This application is for educational purposes only and should not be considered a medical diagnosis. Always consult a healthcare professional.

🎯 Key Learning Outcomes

Handling imbalanced datasets

Hyperparameter tuning using GridSearchCV

Evaluating ML models beyond accuracy

Saving & loading production models

Deploying ML applications to the cloud

Building interactive ML dashboards

👨‍💻 Author

Rakhal Krishna AI & Data Science Student

GitHub: https://github.com/Rakhal06

---

🚀 Run Locally

git clone https://github.com/Rakhal06/Diabetes-Prediction-System.git
cd Diabetes-Prediction-System
python -m venv venv
venv\Scripts\activate   # Windows
pip install -r requirements.txt
streamlit run app.py