🔬 PCOS Research & Risk Prediction using Machine Learning

A research-focused project aimed at analyzing clinical data of PCOS (Polycystic Ovary Syndrome) patients to identify critical features, derive meaningful insights, and build predictive ML models that help in early detection and better healthcare decision-making.

---

⭐ Project Objectives

• Understand PCOS prevalence through clinical and lifestyle features
• Perform extensive Exploratory Data Analysis (EDA)
• Develop ML-based PCOS Risk Prediction Models
• Evaluate model performance and feature significance
• Support women’s health research through data-driven insights

---

📂 Project Structure

📁 PCOS-Research
├─ 📄 data.csv                # Dataset used for analysis
├─ 📓 EDA_V1.ipynb            # Exploratory Data Analysis notebook
├─ 📓 model.ipynb             # Model building & evaluation notebook
├─ 📄 requirements.txt        # Dependencies
└─ 📁 venv                    # Python virtual environment

---

🧠 Techniques & Methods

Task	Tools / Libraries
EDA	Pandas, NumPy, Matplotlib, Seaborn
Preprocessing	Missing Values, Scaling
ML Models	Logistic Regression, Random Forest, SVM
Evaluation	Confusion Matrix, Accuracy, ROC-AUC

---

📊 Insights Covered

• Feature correlation with PCOS occurrence
• Clinical markers with highest predictive strength
• Distribution differences in patients vs non-patients
• Risk classification using ML models

(📌 Visualizations included inside notebooks)

---

🚀 How to Run

# Clone repository
git clone https://github.com/amitabh-7t/PCOS-Reserch

cd PCOS-Reserch

# Create & activate virtual environment (optional)
python -m venv venv
source venv/bin/activate   # Mac/Linux
venv\Scripts\activate.bat  # Windows

# Install dependencies
pip install -r requirements.txt

# Open notebooks
jupyter notebook

---

🔮 Future Enhancements

• Deploy model as a web-based diagnostic tool
• Add Explainable AI (SHAP) for clinical interpretation
• Expand dataset – more patient diversity
• Integrate real medical reports & ultrasound markers
• Research publication in healthcare AI domain

---

⚠ Research Disclaimer

This project is dedicated to learning & research purposes only. It must not be used for real medical diagnosis without clinical validation.

---

👤 Author

Amitabh Thakur B.Tech CSE (AI & ML) | Healthcare AI Researcher Dayananda Sagar University, Bangalore Founder — Humans Care Foundation

📌 Focus: AI for Women’s Health & Preventive Medicine