I am Kushar Raj Kashyap, a final-year Master of Computer Applications (MCA) student at Haldia Institute of Technology, West Bengal. This project, Disaster Recovery Advisor, was developed as part of my major project submission for the 4th semester (2024–2025), under the guidance of Ms. Kankana Datta, Assistant Professor, Department of Computer Applications.
An AI/ML-powered project designed to predict natural disasters and assist in efficient post-disaster recovery planning using intelligent resource allocation and recommendation systems.
The Disaster Recovery Advisor is an AI/ML-based system that predicts natural disasters such as floods, earthquakes, and storms using historical data, geographical attributes (latitude, longitude), and time-based environmental factors. The platform provides not only disaster prediction but also offers impact estimation (casualties, economic loss, infrastructure damage) and tailored recovery strategies like evacuation plans, medical assistance, and relief inventory management.
- Predict type, severity, and duration of upcoming natural disasters.
- Generate custom recovery plans and smart resource distribution strategies.
- Assist government bodies, NGOs, and citizens in real-time disaster preparedness.
- Monitor and manage disaster relief inventory efficiently.
| Category | Technology/Tools |
|---|---|
| Programming | Python 3.x |
| ML Frameworks | Scikit-learn, Pandas, NumPy |
| Backend | Flask / FastAPI |
| Frontend (optional) | HTML, CSS, JavaScript |
| Visualization | Matplotlib, Seaborn |
| Database | SQLite |
| Dev Tools | VS Code, Jupyter, GitHub |
- ✅ AI-Powered Disaster Prediction using Random Forest
- 📈 Impact Forecasting (e.g., disaster duration, casualty estimation)
- 🚨 Real-Time Recovery Suggestions
- 🚑 Relief Resource Monitoring System
- 🧾 User Roles for General Public, Government, and NGOs
- 📊 Interactive Dashboard with visual insights
- Presentation Layer: User Interface (Login, Dashboard)
- Business Logic Layer: Receives user input, triggers ML processes
- Application Layer: Prediction engine + Recommendation system
- Data Layer: Historical disaster dataset + SQLite DB
| Feature | Expected Result | Status |
|---|---|---|
| Disaster Type Prediction | Predicts correctly with >85% F1 score | ✅ |
| Relief Material Planner | Warns if insufficient inventory | ✅ |
| Recovery Plan Suggestion | Suggests valid recovery strategies | ✅ |
| Invalid Input Handling | Graceful message for unknown disaster types | ✅ |
- Clone the repository
git clone https://github.com/yourusername/disaster-recovery-advisor.git cd disaster-recovery-advisor