🩺 Pregnancy Health Risk Level Prediction (ML Project)

This project involves predicting the Risk Level of pregnant women based on various health parameters using supervised machine learning models. The goal is to assist healthcare providers in identifying high-risk pregnancies using data-driven insights.

---

📊 Dataset Overview

The dataset contains the following features:

• Age
• Systolic BP
• Diastolic BP
• Blood Sugar (BS)
• Body Temperature
• BMI
• Previous Complications
• Preexisting Diabetes
• Gestational Diabetes
• Mental Health
• Heart Rate
• Risk Level (Target Variable)

---

🧹 Data Preprocessing

• Missing values handled using KNN Imputer for numeric features.
• Outliers treated using Z-score (threshold ±3).
• Categorical variables label encoded.
• Exploratory Data Analysis (EDA) and visualizations performed using seaborn and matplotlib.

---

🤖 Machine Learning Models Used

• Logistic Regression
• Decision Tree Classifier
• Random Forest Classifier
• K-Nearest Neighbors
• Support Vector Machine
• (Optional) XGBoost (Tried but not included due to compatibility issues)

Model Evaluation Metrics:

• Accuracy
• Classification Report (Precision, Recall, F1-score)
• Confusion Matrix

---

✅ Best Performing Model

• Random Forest Classifier achieved the highest accuracy and balanced performance across all classes.
• Hyperparameter tuning can be performed for further optimization.

---

📌 Key Insights

• Features like Previous Complications, BMI, and Blood Sugar showed strong correlation with Risk Level.
• KNN Imputer preserved data trends better than mean/median imputation.
• Random Forest provided better generalization and handled non-linearities effectively.

---

📁 Folder Structure

📦PregnancyRiskPrediction
┣ 📜README.md
┣ 📜data.csv
┣ 📜notebook.ipynb
┗ 📜requirements.txt

---

🚀 How to Run

git clone https://github.com/your-username/PregnancyRiskPrediction.git
cd PregnancyRiskPrediction
pip install -r requirements.txt

Open the notebook.ipynb file and run the cells in order.

---

📬 Contact

Created with ❤️ by Kumari Shivangi 📧 Email: info.kumarishivangi@gmail.com 🔗 LinkedIn