4Dynamics| A Flood of Ideas
A Flood of Ideas
Flood Prediction and Management System
Summary
The program will analyse the live data of rainfall and other weather components. It will raise a flag when a chance of flood is detected. By this way we can predict the upcoming floods and aware the Government as well as NGO’s for a rescue or a excavation mission. The data of the Rainfall, flood, weather conditions, etc. will be then stored and shown on a dashboard for future analysis.As soon as, a pattern related to a flood is detected, a `flag` will be raised and a bunch of notifications will be triggered .Kibana dashboards will be used for visualization, to ensure fast and easy deployment and low maintenance.We are using the Root Mean Square Error for the prediction system.
How We Addressed This Challenge
We developed CARITAS (which means "Lifesaver"). It will send alerts as flags whenever any chances of flood is there. We mean to increase the accuracy of our program by using the Root Mean Square Error.
It is very important as it predicts floods. There was a lot of loss of lives due to floods each year in India and abroad. The flood victims and their families undergo a difficult moment during the floods as they destroy their houses, force them to excavate to other areas and cause suffering everywhere.
We are trying to make a product or software for humanity, that can save lives. This product will be marketed for different government agencies for the safety of citizens.
We also look forward to make system for predicting not only floods, but also other calamities and disasters which cause huge damage on all over the Earth.
How We Developed This Project
The Inspiration
Flood is a very common issues in our locality as well as in most part of India and even the world.
Almost an average of 508 people die due to floods in India, each year. Many of them lose their houses, cannot work, face hunger and thirst issues, etc. That is why we chose this challenge to make the lives of the people easy by predicting the floods beforehand.
The Problems
We have faced a problem of collecting live data and embed them in our code,
The Achievements
We are here to the national's, and we have learned more how industries work, business models, and more about machine learning too.
Coding Language and Software
The coding language we used is Python, and the software we used is Google Colab.
How We Used Space Agency Data in This Project
We are using Malawi datasets (opensource) as the datasets for the preliminary model, once the preliminary model is trained, Merra-II (NASA GISS) will be used as the base for the transfer learning. Depending on the error value GPCP 1-degree-Dail y(NASA GISS) could be used instead of the Mera-II to get better results.
Project Demo
Data & Resources
Malawi,
Merra-II (NASA GISS),
Kibana,
Google Colab,
GPCP 1-degree-Dail y(NASA GISS)