Predictive Pulse: Harnessing Machine Learning for Blood Pressure Analysis

Predictive Pulse is a comprehensive healthcare-focused machine learning project designed to classify hypertension stages based on patient physiological data. By leveraging data science workflows—from Exploratory Data Analysis (EDA) to model deployment—this project provides a user-friendly tool for healthcare providers to monitor and manage patient blood pressure effectively.

---

🚀 Key Features

• Data Analysis & Preprocessing: In-depth cleaning and pattern recognition using Pandas and NumPy.
• Visual Exploration: Detailed EDA and learning curve visualizations using Matplotlib and Seaborn.
• Multiple Model Training: Comparison of Logistic Regression, Decision Trees, Random Forest, and SVM (Support Vector Classifier).
• Persistent Model Storage: Trained models are serialized using joblib for efficient reuse in production.
• Interactive Web Application: A Flask-based dashboard allowing users to input patient data and receive real-time hypertension predictions.

---

🛠️ Tech Stack

• Language: Python 3.12
• Libraries: Scikit-learn, Pandas, NumPy
• Visualization: Matplotlib, Seaborn
• Web Framework: Flask
• Environment: Jupyter Notebook

---

💻 Installation and Setup

1. Environment Setup

Clone the repository and create a virtual environment to keep dependencies isolated:

git clone <your-repository-url>
cd Predictive-Pulse
python -m venv venv
# Activate on Windows:
.\venv\Scripts\activate

2. Install Dependencies

pip install -r requirements.txt
pip install notebook matplotlib seaborn  # To ensure notebook tools are available

3. Model Training

Before running the app, the model must be trained using the provided dataset:

python -m notebook model.ipynb

• In the Jupyter interface, go to Cell > Run All.
• This will generate the model.joblib file required for predictions.

4. Run the Web Application

python app.py

Visit http://127.0.0.1:5000/ in your browser to interact with the project.

---

📊 Methodology

1. Data Preprocessing: Handling missing values and scaling numerical features using StandardScaler.
2. Model Selection: Testing various classifiers to find the optimal balance between precision and recall.
3. Deployment: Integrating the trained SVC (Support Vector Classifier) into a Flask backend to serve predictions to a frontend interface.

---

📝 Internship Details

• Project Name: Predictive Pulse
• Submission Date: December 17, 2025
• Developed by: SHreenath Tambe

---

How to use this:

1. Open the README.md file in your project folder using Notepad or VS Code.
2. Delete everything currently inside it.
3. Paste the text above.
4. Save it.
5. Commit and push to GitHub:

git add README.md
git commit -m "Update README with professional documentation"
git push origin main