Singapore-India Hackathon 2019 Epidemiological Prediction using RNNs (LSTM, RNN, GRU).
Presently, centralized data repository is missing for various diseases and citizens for different regions of the country. The solution required a predictive analysis and medication support using epidemiology of citizens, diseases, and geography. The application should be able to assist policy makers provide better health services to affected areas. More information about the problem statement & competition here.
We prototyped a dashboard for policy makers that uses RNNs to predict future disease hotspots, optimization solvers to suggest movement of healthcare supplies. We also developed an Android app for quick reporting of doctors on the ground, using QR scanning of the Indian ID, and text-to-speech reporting of diagnoses.
Check out our final pitch deck and play around with our RNN predictive model! We included more technical details of the RNN predictive model here.
The amazing experience I had in India wouldn't be the same without my team members and friends Darren, Yong Keat, Rohit, Gargi and Samarth. Our mentors Wayne and. Aditya had been tremendously helpful in shaping our solution and crafting our pitch direction, and the entire event took off without a hitch only because of the organizing teams behind it. Thank you IIT Madras for hosting us!