taapasx28 / codefundoo_project Goto Github PK

We are planning to build and API that would be focussing on predicting and managing the natural disaster thunderstorms.

We have used a dataset which has weather conditions of the city Austin on a daily basis and with adverse weather condition if prevalent on that day. The dataset was found on the site Kaggle (https://www.kaggle.com/grubenm/austin-weather).We shifted the column of adverse weather event upwards by one cell, so now today’s average weather condition is mapped with tomorrow’s adverse weather event. And now on this dataset we trained our predictive model.And we came up with some interesting results.

We used Azure ML Studio to train our model.

And since we have to implement it in areas with internet connectivity issues, we will use colour codes which will describe the danger level of the situation(i.e time left before the outburst) to notify people as well as the panchayat of each village area or the municipal corporation of the town , so that they can help in proper reallocation of people within stipulated time. Colour codes will ensure that even an illiterate person can realize the possible threat. This is the basic idea for our codefundo++ project.

Color codes will be 3 i.e Red,Yellow and Green . Green symbolizes no danger at present. Yellow suggests moderate level of dangers suggesting people to pack stuff and move out . Red will symbolize condition of absolute danger where going of to rescue camps will be the the best option.

We used Flask API alond with Microsoft Aure ML studio API key to present our ideas as a form which can take input the present day data and states the current situation.

Link to our slide description (https://docs.google.com/presentation/d/19hIj4oPox4vm0yEkozE5twub3RP3EOnZiCPpZrhR-uM/edit?usp=sharing)

Run application.py for execution.

Run localhost ip then indicated on your browser.