💊 Drug Prediction using Machine Learning (Django)

A Django-based web application that predicts the most suitable drug for a patient based on their medical information. Powered by a trained machine learning model, this app demonstrates the real-world application of AI in healthcare.

---

📸 Interface Preview Demo of prediction form:

---

🚀 Features

• 🧾 Patient Data Form: User-friendly form to collect:
  • Age
  • Gender (Sex)
  • Blood Pressure
  • Cholesterol
  • Sodium-to-Potassium Ratio (Na_to_K)
• 🤖 ML Drug Prediction: Predicts the most effective drug using a pre-trained ML model.
• 🔐 Admin Panel: Manage entries and oversee patient predictions with Django’s built-in admin interface.
• 📈 Model Integration: Seamlessly integrates a trained scikit-learn model into the Django backend.
• 🎯 Accurate Predictions: Based on well-curated training data and effective preprocessing.

---

🧠 Machine Learning Model

• Algorithm Used: (e.g., Decision Tree Classifier / Random Forest / Logistic Regression)
• Training Data: [Include dataset source or summary if applicable]
• Target Classes: Drug A, Drug B, Drug C, Drug X, Drug Y (example)
• Input Features:
  • Age (Integer)
  • Sex (Male/Female)
  • Blood Pressure (Low/Normal/High)
  • Cholesterol (Normal/High)
  • Na_to_K (Float – Sodium-to-Potassium Ratio)

---

🖥️ Tech Stack

• Backend: Django (Python)
• Machine Learning: scikit-learn, pandas, joblib
• Frontend: HTML5, CSS3, Bootstrap (Optional)
• Database: SQLite3 (default Django DB)
• Model Deployment: Integrated with Django views

---

⚙️ Setup Instructions

1. Clone the Repository

git clone https://github.com/your-username/drug-prediction-django.git
cd drug-prediction-django

2. Install Dependencies

pip install -r requirements.txt

3. Apply Migrations

python manage.py migrate

4. Run the Development Server

python manage.py runserver

5. Access the App

Open your browser and go to:

http://127.0.0.1:8000/

---

🧪 Model Training (Optional for Developers)

If you want to retrain or improve the model:

# train_model.py
# Uses scikit-learn to train a model and saves it as 'model.pkl' using joblib

---

📸 Screenshots

Add screenshots of your:

• 🧾 Patient input form
• 📈 Prediction result page
• 🔐 Admin panel view

---

📁 Project Structure

drug-prediction/
├── core/                 # Django app
│   ├── models.py         # Patient model
│   ├── views.py          # Prediction logic
│   ├── forms.py          # Input form
│   └── templates/        # HTML templates
├── static/               # CSS/JS assets
├── model/                # Trained ML model (model.pkl)
├── train_model.py        # (Optional) Script to train and save model
├── db.sqlite3
├── manage.py
└── requirements.txt

---

✅ Example Input & Output

Input Field	Sample Value
Age	35
Gender	Male
Blood Pressure	High
Cholesterol	High
Na_to_K	14.2

Predicted Drug: DrugX

---

🔐 Admin Credentials

To access the admin panel:

python manage.py createsuperuser

Then visit:

http://127.0.0.1:8000/admin

---

📜 License

MIT License. Feel free to use, modify, and distribute this project.

---

🙌 Acknowledgements

• scikit-learn
• Django
• UCI Drug Dataset (if applicable)
• Bootstrap (for styling)

---