In our daily life we always see so many natural disasters occurs in day to day life and we don't get its prior notification or warning due to which we become the victim of those disasters. So why not to built a system which can predict those disasters and warn us.
What it does
It predicts the future disasters based on the machined trained from the historical data of disasters. It will predict the future disaster based on the past data and will send push notification to the users with the location to nearest safe place and time remaining for the disasters to come.
How we built it
We have used data sets of previous flood occurrence in few cities of India and based on those data we have trained the model. For training the machine and future prediction we used Azure ML Studio. For fetching the data from apis we used python script and gave the data to Machine for future prediction. For users we parsed the data to web and android app so that he can stay updated with the recent weather changes and upcoming disasters.
Challenges we ran into
Getting the data sets Connecting APIs data to ML Studio
Accomplishments that we're proud of
We trained the model which can give the precision output up to 0.998
What we learned
Machine Learning, Push Notification
What's next for Intelligent Disaster Prediction and Prevention System
Collecting more accurate data for training machine Adding up more features like:
- Showing up the help arriving for the victims
- Locating the victims location and asking for the peoples to help
- Training the model for more natural disasters.